diff options
Diffstat (limited to 'src/Norm4Functions.c')
| -rw-r--r-- | src/Norm4Functions.c | 852 |
1 files changed, 594 insertions, 258 deletions
diff --git a/src/Norm4Functions.c b/src/Norm4Functions.c index f106642..795bd23 100644 --- a/src/Norm4Functions.c +++ b/src/Norm4Functions.c @@ -18,13 +18,18 @@ extern "C" { #endif -int LocalReduce_Norm4_BYTE(int i, int mask_on, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[]) +int LocalReduce_Norm4_BYTE(int i, int weight_on, const void * const weight, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[], int param_table_handle) { int iter = 0; int sum_indices = 0; - int flag, product, j, k; + int flag, product, j, k, ierr; + + /* Weight variables */ + CCTK_REAL weight_sum = 0.0; + CCTK_REAL weight_value = 1.0; #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -36,6 +41,7 @@ int LocalReduce_Norm4_BYTE(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -43,51 +49,52 @@ int LocalReduce_Norm4_BYTE(int i, int mask_on, CCTK_INT * input_array_offsets, i { /* out values type switches*/ case CCTK_VARIABLE_BYTE: - ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_BYTE, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_BYTE, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; case CCTK_VARIABLE_INT: - ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_INT, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_INT, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_INT1 case CCTK_VARIABLE_INT1: - ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_INT1, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_INT1, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT2 case CCTK_VARIABLE_INT2: - ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_INT2, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_INT2, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT4 case CCTK_VARIABLE_INT4: - ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_INT4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_INT4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT8 case CCTK_VARIABLE_INT8: - ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_INT8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_INT8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_REAL: - ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_REAL, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_REAL, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_REAL4 case CCTK_VARIABLE_REAL4: - ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_REAL4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_REAL4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL8 case CCTK_VARIABLE_REAL8: - ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_REAL8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_REAL8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL16 case CCTK_VARIABLE_REAL16: - ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_REAL16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_REAL16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_COMPLEX: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -111,10 +118,12 @@ int LocalReduce_Norm4_BYTE(int i, int mask_on, CCTK_INT * input_array_offsets, i #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_COMPLEX, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_COMPLEX, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -126,6 +135,7 @@ int LocalReduce_Norm4_BYTE(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -133,6 +143,7 @@ int LocalReduce_Norm4_BYTE(int i, int mask_on, CCTK_INT * input_array_offsets, i #ifdef CCTK_COMPLEX8 case CCTK_VARIABLE_COMPLEX8: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -155,10 +166,12 @@ int LocalReduce_Norm4_BYTE(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd8( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd8( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -170,6 +183,7 @@ int LocalReduce_Norm4_BYTE(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -178,6 +192,7 @@ int LocalReduce_Norm4_BYTE(int i, int mask_on, CCTK_INT * input_array_offsets, i #ifdef CCTK_COMPLEX16 case CCTK_VARIABLE_COMPLEX16: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -200,11 +215,13 @@ int LocalReduce_Norm4_BYTE(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -216,6 +233,7 @@ int LocalReduce_Norm4_BYTE(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -224,6 +242,7 @@ int LocalReduce_Norm4_BYTE(int i, int mask_on, CCTK_INT * input_array_offsets, i #ifdef CCTK_COMPLEX32 case CCTK_VARIABLE_COMPLEX32: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -246,11 +265,13 @@ int LocalReduce_Norm4_BYTE(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_BYTE, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -262,21 +283,29 @@ int LocalReduce_Norm4_BYTE(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) break; #endif } - return num_points; + ierr = Util_TableSetInt(param_table_handle, num_points, "num_points"); + ierr = Util_TableSetReal(param_table_handle, weight_sum, "weight_sum"); + return 0; } -int LocalReduce_Norm4_INT(int i, int mask_on, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[]) +int LocalReduce_Norm4_INT(int i, int weight_on, const void * const weight, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[], int param_table_handle) { int iter = 0; int sum_indices = 0; - int flag, product, j, k; + int flag, product, j, k, ierr; + + /* Weight variables */ + CCTK_REAL weight_sum = 0.0; + CCTK_REAL weight_value = 1.0; #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -288,6 +317,7 @@ int LocalReduce_Norm4_INT(int i, int mask_on, CCTK_INT * input_array_offsets, in #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -295,51 +325,52 @@ int LocalReduce_Norm4_INT(int i, int mask_on, CCTK_INT * input_array_offsets, in { /* out values type switches*/ case CCTK_VARIABLE_BYTE: - ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_BYTE, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_BYTE, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; case CCTK_VARIABLE_INT: - ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_INT, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_INT, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_INT1 case CCTK_VARIABLE_INT1: - ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_INT1, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_INT1, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT2 case CCTK_VARIABLE_INT2: - ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_INT2, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_INT2, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT4 case CCTK_VARIABLE_INT4: - ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_INT4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_INT4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT8 case CCTK_VARIABLE_INT8: - ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_INT8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_INT8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_REAL: - ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_REAL, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_REAL, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_REAL4 case CCTK_VARIABLE_REAL4: - ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_REAL4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_REAL4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL8 case CCTK_VARIABLE_REAL8: - ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_REAL8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_REAL8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL16 case CCTK_VARIABLE_REAL16: - ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_REAL16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_REAL16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_COMPLEX: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -362,10 +393,12 @@ int LocalReduce_Norm4_INT(int i, int mask_on, CCTK_INT * input_array_offsets, in inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_COMPLEX, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_COMPLEX, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -377,6 +410,7 @@ int LocalReduce_Norm4_INT(int i, int mask_on, CCTK_INT * input_array_offsets, in #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -384,6 +418,7 @@ int LocalReduce_Norm4_INT(int i, int mask_on, CCTK_INT * input_array_offsets, in #ifdef CCTK_COMPLEX8 case CCTK_VARIABLE_COMPLEX8: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -406,10 +441,12 @@ int LocalReduce_Norm4_INT(int i, int mask_on, CCTK_INT * input_array_offsets, in inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd8( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd8( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -421,6 +458,7 @@ int LocalReduce_Norm4_INT(int i, int mask_on, CCTK_INT * input_array_offsets, in #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -429,6 +467,7 @@ int LocalReduce_Norm4_INT(int i, int mask_on, CCTK_INT * input_array_offsets, in #ifdef CCTK_COMPLEX16 case CCTK_VARIABLE_COMPLEX16: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -451,10 +490,12 @@ int LocalReduce_Norm4_INT(int i, int mask_on, CCTK_INT * input_array_offsets, in inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -466,6 +507,7 @@ int LocalReduce_Norm4_INT(int i, int mask_on, CCTK_INT * input_array_offsets, in #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -474,6 +516,7 @@ int LocalReduce_Norm4_INT(int i, int mask_on, CCTK_INT * input_array_offsets, in #ifdef CCTK_COMPLEX32 case CCTK_VARIABLE_COMPLEX32: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -496,10 +539,12 @@ int LocalReduce_Norm4_INT(int i, int mask_on, CCTK_INT * input_array_offsets, in inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -511,22 +556,30 @@ int LocalReduce_Norm4_INT(int i, int mask_on, CCTK_INT * input_array_offsets, in #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) break; #endif } - return num_points; + ierr = Util_TableSetInt(param_table_handle, num_points, "num_points"); + ierr = Util_TableSetReal(param_table_handle, weight_sum, "weight_sum"); + return 0; } #ifdef CCTK_INT1 -int LocalReduce_Norm4_INT1(int i, int mask_on, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[]) +int LocalReduce_Norm4_INT1(int i, int weight_on, const void * const weight, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[], int param_table_handle) { int iter = 0; int sum_indices = 0; - int flag, product, j, k; + int flag, product, j, k, ierr; + + /* Weight variables */ + CCTK_REAL weight_sum = 0.0; + CCTK_REAL weight_value = 1.0; #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -538,6 +591,7 @@ int LocalReduce_Norm4_INT1(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -545,51 +599,52 @@ int LocalReduce_Norm4_INT1(int i, int mask_on, CCTK_INT * input_array_offsets, i { /* out values type switches*/ case CCTK_VARIABLE_BYTE: - ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_BYTE, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_BYTE, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; case CCTK_VARIABLE_INT: - ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_INT, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_INT, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_INT11 case CCTK_VARIABLE_INT1: - ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_INT1, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_INT1, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT12 case CCTK_VARIABLE_INT2: - ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_INT2, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_INT2, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT14 case CCTK_VARIABLE_INT4: - ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_INT4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_INT4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT18 case CCTK_VARIABLE_INT8: - ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_INT8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_INT8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_REAL: - ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_REAL, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_REAL, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_REAL4 case CCTK_VARIABLE_REAL4: - ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_REAL4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_REAL4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL8 case CCTK_VARIABLE_REAL8: - ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_REAL8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_REAL8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL16 case CCTK_VARIABLE_REAL16: - ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_REAL16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_REAL16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_COMPLEX: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -612,10 +667,12 @@ int LocalReduce_Norm4_INT1(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_COMPLEX, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_COMPLEX, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -627,6 +684,7 @@ int LocalReduce_Norm4_INT1(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -634,6 +692,7 @@ int LocalReduce_Norm4_INT1(int i, int mask_on, CCTK_INT * input_array_offsets, i #ifdef CCTK_COMPLEX8 case CCTK_VARIABLE_COMPLEX8: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -656,10 +715,12 @@ int LocalReduce_Norm4_INT1(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd8( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd8( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -671,6 +732,7 @@ int LocalReduce_Norm4_INT1(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -679,6 +741,7 @@ int LocalReduce_Norm4_INT1(int i, int mask_on, CCTK_INT * input_array_offsets, i #ifdef CCTK_COMPLEX16 case CCTK_VARIABLE_COMPLEX16: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -701,11 +764,13 @@ int LocalReduce_Norm4_INT1(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -717,6 +782,7 @@ int LocalReduce_Norm4_INT1(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -725,6 +791,7 @@ int LocalReduce_Norm4_INT1(int i, int mask_on, CCTK_INT * input_array_offsets, i #ifdef CCTK_COMPLEX32 case CCTK_VARIABLE_COMPLEX32: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -747,11 +814,13 @@ int LocalReduce_Norm4_INT1(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT1, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -763,23 +832,31 @@ int LocalReduce_Norm4_INT1(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) break; #endif } - return num_points; + ierr = Util_TableSetInt(param_table_handle, num_points, "num_points"); + ierr = Util_TableSetReal(param_table_handle, weight_sum, "weight_sum"); + return 0; } #endif #ifdef CCTK_INT2 -int LocalReduce_Norm4_INT2(int i, int mask_on, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[]) +int LocalReduce_Norm4_INT2(int i, int weight_on, const void * const weight, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[], int param_table_handle) { int iter = 0; int sum_indices = 0; - int flag, product, j, k; + int flag, product, j, k, ierr; + + /* Weight variables */ + CCTK_REAL weight_sum = 0.0; + CCTK_REAL weight_value = 1.0; #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -791,6 +868,7 @@ int LocalReduce_Norm4_INT2(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -798,51 +876,52 @@ int LocalReduce_Norm4_INT2(int i, int mask_on, CCTK_INT * input_array_offsets, i { /* out values type switches*/ case CCTK_VARIABLE_BYTE: - ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_BYTE, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_BYTE, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; case CCTK_VARIABLE_INT: - ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_INT, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_INT, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_INT21 case CCTK_VARIABLE_INT1: - ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_INT1, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_INT1, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT22 case CCTK_VARIABLE_INT2: - ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_INT2, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_INT2, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT24 case CCTK_VARIABLE_INT4: - ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_INT4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_INT4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT28 case CCTK_VARIABLE_INT8: - ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_INT8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_INT8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_REAL: - ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_REAL, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_REAL, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_REAL4 case CCTK_VARIABLE_REAL4: - ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_REAL4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_REAL4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL8 case CCTK_VARIABLE_REAL8: - ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_REAL8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_REAL8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL16 case CCTK_VARIABLE_REAL16: - ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_REAL16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_REAL16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_COMPLEX: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -865,10 +944,12 @@ int LocalReduce_Norm4_INT2(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_COMPLEX, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_COMPLEX, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -880,6 +961,7 @@ int LocalReduce_Norm4_INT2(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -887,6 +969,7 @@ int LocalReduce_Norm4_INT2(int i, int mask_on, CCTK_INT * input_array_offsets, i #ifdef CCTK_COMPLEX8 case CCTK_VARIABLE_COMPLEX8: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -909,10 +992,12 @@ int LocalReduce_Norm4_INT2(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd8( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd8( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -924,6 +1009,7 @@ int LocalReduce_Norm4_INT2(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -932,6 +1018,7 @@ int LocalReduce_Norm4_INT2(int i, int mask_on, CCTK_INT * input_array_offsets, i #ifdef CCTK_COMPLEX16 case CCTK_VARIABLE_COMPLEX16: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -954,11 +1041,13 @@ int LocalReduce_Norm4_INT2(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -970,6 +1059,7 @@ int LocalReduce_Norm4_INT2(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -978,6 +1068,7 @@ int LocalReduce_Norm4_INT2(int i, int mask_on, CCTK_INT * input_array_offsets, i #ifdef CCTK_COMPLEX32 case CCTK_VARIABLE_COMPLEX32: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1000,11 +1091,13 @@ int LocalReduce_Norm4_INT2(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT2, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1016,23 +1109,31 @@ int LocalReduce_Norm4_INT2(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) break; #endif } - return num_points; + ierr = Util_TableSetInt(param_table_handle, num_points, "num_points"); + ierr = Util_TableSetReal(param_table_handle, weight_sum, "weight_sum"); + return 0; } #endif #ifdef CCTK_INT4 -int LocalReduce_Norm4_INT4(int i, int mask_on, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[]) +int LocalReduce_Norm4_INT4(int i, int weight_on, const void * const weight, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[], int param_table_handle) { int iter = 0; int sum_indices = 0; - int flag, product, j, k; + int flag, product, j, k, ierr; + + /* Weight variables */ + CCTK_REAL weight_sum = 0.0; + CCTK_REAL weight_value = 1.0; #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1044,6 +1145,7 @@ int LocalReduce_Norm4_INT4(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -1051,51 +1153,52 @@ int LocalReduce_Norm4_INT4(int i, int mask_on, CCTK_INT * input_array_offsets, i { /* out values type switches*/ case CCTK_VARIABLE_BYTE: - ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_BYTE, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_BYTE, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; case CCTK_VARIABLE_INT: - ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_INT, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_INT, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_INT1 case CCTK_VARIABLE_INT1: - ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_INT1, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_INT1, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT2 case CCTK_VARIABLE_INT2: - ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_INT2, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_INT2, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT4 case CCTK_VARIABLE_INT4: - ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_INT4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_INT4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT8 case CCTK_VARIABLE_INT8: - ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_INT8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_INT8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_REAL: - ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_REAL, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_REAL, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_REAL4 case CCTK_VARIABLE_REAL4: - ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_REAL4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_REAL4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL8 case CCTK_VARIABLE_REAL8: - ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_REAL8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_REAL8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL16 case CCTK_VARIABLE_REAL16: - ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_REAL16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_REAL16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_COMPLEX: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1118,10 +1221,12 @@ int LocalReduce_Norm4_INT4(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_COMPLEX, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_COMPLEX, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1133,6 +1238,7 @@ int LocalReduce_Norm4_INT4(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -1140,6 +1246,7 @@ int LocalReduce_Norm4_INT4(int i, int mask_on, CCTK_INT * input_array_offsets, i #ifdef CCTK_COMPLEX8 case CCTK_VARIABLE_COMPLEX8: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1162,10 +1269,12 @@ int LocalReduce_Norm4_INT4(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd8( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd8( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1177,6 +1286,7 @@ int LocalReduce_Norm4_INT4(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -1185,6 +1295,7 @@ int LocalReduce_Norm4_INT4(int i, int mask_on, CCTK_INT * input_array_offsets, i #ifdef CCTK_COMPLEX16 case CCTK_VARIABLE_COMPLEX16: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1207,11 +1318,13 @@ int LocalReduce_Norm4_INT4(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1223,6 +1336,7 @@ int LocalReduce_Norm4_INT4(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -1231,6 +1345,7 @@ int LocalReduce_Norm4_INT4(int i, int mask_on, CCTK_INT * input_array_offsets, i #ifdef CCTK_COMPLEX32 case CCTK_VARIABLE_COMPLEX32: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1253,11 +1368,13 @@ int LocalReduce_Norm4_INT4(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT4, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1269,23 +1386,31 @@ int LocalReduce_Norm4_INT4(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) break; #endif } - return num_points; + ierr = Util_TableSetInt(param_table_handle, num_points, "num_points"); + ierr = Util_TableSetReal(param_table_handle, weight_sum, "weight_sum"); + return 0; } #endif #ifdef CCTK_INT8 -int LocalReduce_Norm4_INT8(int i, int mask_on, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[]) +int LocalReduce_Norm4_INT8(int i, int weight_on, const void * const weight, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[], int param_table_handle) { int iter = 0; int sum_indices = 0; - int flag, product, j, k; + int flag, product, j, k, ierr; + + /* Weight variables */ + CCTK_REAL weight_sum = 0.0; + CCTK_REAL weight_value = 1.0; #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1297,6 +1422,7 @@ int LocalReduce_Norm4_INT8(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -1304,51 +1430,52 @@ int LocalReduce_Norm4_INT8(int i, int mask_on, CCTK_INT * input_array_offsets, i { /* out values type switches*/ case CCTK_VARIABLE_BYTE: - ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_BYTE, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_BYTE, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; case CCTK_VARIABLE_INT: - ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_INT, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_INT, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_INT1 case CCTK_VARIABLE_INT1: - ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_INT1, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_INT1, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT2 case CCTK_VARIABLE_INT2: - ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_INT2, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_INT2, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT4 case CCTK_VARIABLE_INT4: - ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_INT4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_INT4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT8 case CCTK_VARIABLE_INT8: - ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_INT8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_INT8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_REAL: - ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_REAL, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_REAL, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_REAL4 case CCTK_VARIABLE_REAL4: - ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_REAL4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_REAL4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL8 case CCTK_VARIABLE_REAL8: - ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_REAL8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_REAL8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL16 case CCTK_VARIABLE_REAL16: - ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_REAL16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_REAL16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_COMPLEX: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1371,10 +1498,12 @@ int LocalReduce_Norm4_INT8(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_COMPLEX, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_COMPLEX, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1386,6 +1515,7 @@ int LocalReduce_Norm4_INT8(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -1393,6 +1523,7 @@ int LocalReduce_Norm4_INT8(int i, int mask_on, CCTK_INT * input_array_offsets, i #ifdef CCTK_COMPLEX8 case CCTK_VARIABLE_COMPLEX8: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1415,10 +1546,12 @@ int LocalReduce_Norm4_INT8(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd8( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd8( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1430,6 +1563,7 @@ int LocalReduce_Norm4_INT8(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -1438,6 +1572,7 @@ int LocalReduce_Norm4_INT8(int i, int mask_on, CCTK_INT * input_array_offsets, i #ifdef CCTK_COMPLEX16 case CCTK_VARIABLE_COMPLEX16: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1460,11 +1595,13 @@ int LocalReduce_Norm4_INT8(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1476,6 +1613,7 @@ int LocalReduce_Norm4_INT8(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -1484,6 +1622,7 @@ int LocalReduce_Norm4_INT8(int i, int mask_on, CCTK_INT * input_array_offsets, i #ifdef CCTK_COMPLEX32 case CCTK_VARIABLE_COMPLEX32: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1506,11 +1645,13 @@ int LocalReduce_Norm4_INT8(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_INT8, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1522,22 +1663,30 @@ int LocalReduce_Norm4_INT8(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) break; #endif } - return num_points; + ierr = Util_TableSetInt(param_table_handle, num_points, "num_points"); + ierr = Util_TableSetReal(param_table_handle, weight_sum, "weight_sum"); + return 0; } #endif -int LocalReduce_Norm4_REAL(int i, int mask_on, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[]) +int LocalReduce_Norm4_REAL(int i, int weight_on, const void * const weight, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[], int param_table_handle) { int iter = 0; int sum_indices = 0; - int flag, product, j, k; + int flag, product, j, k, ierr; + + /* Weight variables */ + CCTK_REAL weight_sum = 0.0; + CCTK_REAL weight_value = 1.0; #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1549,6 +1698,7 @@ int LocalReduce_Norm4_REAL(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -1556,51 +1706,52 @@ int LocalReduce_Norm4_REAL(int i, int mask_on, CCTK_INT * input_array_offsets, i { /* out values type switches*/ case CCTK_VARIABLE_BYTE: - ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_BYTE, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_BYTE, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; case CCTK_VARIABLE_INT: - ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_INT, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_INT, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_INT1 case CCTK_VARIABLE_INT1: - ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_INT1, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_INT1, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT2 case CCTK_VARIABLE_INT2: - ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_INT2, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_INT2, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT4 case CCTK_VARIABLE_INT4: - ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_INT4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_INT4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT8 case CCTK_VARIABLE_INT8: - ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_INT8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_INT8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_REAL: - ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_REAL, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_REAL, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_REAL4 case CCTK_VARIABLE_REAL4: - ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_REAL4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_REAL4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL8 case CCTK_VARIABLE_REAL8: - ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_REAL8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_REAL8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL16 case CCTK_VARIABLE_REAL16: - ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_REAL16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_REAL16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_COMPLEX: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1623,10 +1774,12 @@ int LocalReduce_Norm4_REAL(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_COMPLEX, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_COMPLEX, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1638,6 +1791,7 @@ int LocalReduce_Norm4_REAL(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -1645,6 +1799,7 @@ int LocalReduce_Norm4_REAL(int i, int mask_on, CCTK_INT * input_array_offsets, i #ifdef CCTK_COMPLEX8 case CCTK_VARIABLE_COMPLEX8: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1667,10 +1822,12 @@ int LocalReduce_Norm4_REAL(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd8( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd8( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1682,6 +1839,7 @@ int LocalReduce_Norm4_REAL(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -1690,6 +1848,7 @@ int LocalReduce_Norm4_REAL(int i, int mask_on, CCTK_INT * input_array_offsets, i #ifdef CCTK_COMPLEX16 case CCTK_VARIABLE_COMPLEX16: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1712,11 +1871,13 @@ int LocalReduce_Norm4_REAL(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1728,6 +1889,7 @@ int LocalReduce_Norm4_REAL(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -1736,6 +1898,7 @@ int LocalReduce_Norm4_REAL(int i, int mask_on, CCTK_INT * input_array_offsets, i #ifdef CCTK_COMPLEX32 case CCTK_VARIABLE_COMPLEX32: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1758,11 +1921,13 @@ int LocalReduce_Norm4_REAL(int i, int mask_on, CCTK_INT * input_array_offsets, i inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1774,22 +1939,30 @@ int LocalReduce_Norm4_REAL(int i, int mask_on, CCTK_INT * input_array_offsets, i #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) break; #endif } - return num_points; + ierr = Util_TableSetInt(param_table_handle, num_points, "num_points"); + ierr = Util_TableSetReal(param_table_handle, weight_sum, "weight_sum"); + return 0; } #ifdef CCTK_REAL4 -int LocalReduce_Norm4_REAL4(int i, int mask_on, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[]) +int LocalReduce_Norm4_REAL4(int i, int weight_on, const void * const weight, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[], int param_table_handle) { int iter = 0; int sum_indices = 0; - int flag, product, j, k; + int flag, product, j, k, ierr; + + /* Weight variables */ + CCTK_REAL weight_sum = 0.0; + CCTK_REAL weight_value = 1.0; #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1801,6 +1974,7 @@ int LocalReduce_Norm4_REAL4(int i, int mask_on, CCTK_INT * input_array_offsets, #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -1808,51 +1982,52 @@ int LocalReduce_Norm4_REAL4(int i, int mask_on, CCTK_INT * input_array_offsets, { /* out values type switches*/ case CCTK_VARIABLE_BYTE: - ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_BYTE, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_BYTE, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; case CCTK_VARIABLE_INT: - ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_INT, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_INT, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_INT1 case CCTK_VARIABLE_INT1: - ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_INT1, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_INT1, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT2 case CCTK_VARIABLE_INT2: - ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_INT2, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_INT2, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT4 case CCTK_VARIABLE_INT4: - ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_INT4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_INT4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT8 case CCTK_VARIABLE_INT8: - ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_INT8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_INT8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_REAL: - ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_REAL, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_REAL, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_REAL4 case CCTK_VARIABLE_REAL4: - ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_REAL4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_REAL4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL8 case CCTK_VARIABLE_REAL8: - ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_REAL8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_REAL8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL16 case CCTK_VARIABLE_REAL16: - ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_REAL16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_REAL16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_COMPLEX: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1875,10 +2050,12 @@ int LocalReduce_Norm4_REAL4(int i, int mask_on, CCTK_INT * input_array_offsets, inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_COMPLEX, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_COMPLEX, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1890,6 +2067,7 @@ int LocalReduce_Norm4_REAL4(int i, int mask_on, CCTK_INT * input_array_offsets, #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -1897,6 +2075,7 @@ int LocalReduce_Norm4_REAL4(int i, int mask_on, CCTK_INT * input_array_offsets, #ifdef CCTK_COMPLEX8 case CCTK_VARIABLE_COMPLEX8: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1919,10 +2098,12 @@ int LocalReduce_Norm4_REAL4(int i, int mask_on, CCTK_INT * input_array_offsets, inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd8( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd8( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1934,6 +2115,7 @@ int LocalReduce_Norm4_REAL4(int i, int mask_on, CCTK_INT * input_array_offsets, #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -1942,6 +2124,7 @@ int LocalReduce_Norm4_REAL4(int i, int mask_on, CCTK_INT * input_array_offsets, #ifdef CCTK_COMPLEX16 case CCTK_VARIABLE_COMPLEX16: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1964,11 +2147,13 @@ int LocalReduce_Norm4_REAL4(int i, int mask_on, CCTK_INT * input_array_offsets, inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -1980,6 +2165,7 @@ int LocalReduce_Norm4_REAL4(int i, int mask_on, CCTK_INT * input_array_offsets, #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -1988,6 +2174,7 @@ int LocalReduce_Norm4_REAL4(int i, int mask_on, CCTK_INT * input_array_offsets, #ifdef CCTK_COMPLEX32 case CCTK_VARIABLE_COMPLEX32: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -2010,11 +2197,13 @@ int LocalReduce_Norm4_REAL4(int i, int mask_on, CCTK_INT * input_array_offsets, inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL4, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -2026,23 +2215,31 @@ int LocalReduce_Norm4_REAL4(int i, int mask_on, CCTK_INT * input_array_offsets, #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) break; #endif } - return num_points; + ierr = Util_TableSetInt(param_table_handle, num_points, "num_points"); + ierr = Util_TableSetReal(param_table_handle, weight_sum, "weight_sum"); + return 0; } #endif #ifdef CCTK_REAL8 -int LocalReduce_Norm4_REAL8(int i, int mask_on, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[]) +int LocalReduce_Norm4_REAL8(int i, int weight_on, const void * const weight, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[], int param_table_handle) { int iter = 0; int sum_indices = 0; - int flag, product, j, k; + int flag, product, j, k, ierr; + + /* Weight variables */ + CCTK_REAL weight_sum = 0.0; + CCTK_REAL weight_value = 1.0; #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -2054,6 +2251,7 @@ int LocalReduce_Norm4_REAL8(int i, int mask_on, CCTK_INT * input_array_offsets, #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -2061,51 +2259,52 @@ int LocalReduce_Norm4_REAL8(int i, int mask_on, CCTK_INT * input_array_offsets, { /* out values type switches*/ case CCTK_VARIABLE_BYTE: - ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_BYTE, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_BYTE, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; case CCTK_VARIABLE_INT: - ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_INT, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_INT, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_INT1 case CCTK_VARIABLE_INT1: - ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_INT1, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_INT1, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT2 case CCTK_VARIABLE_INT2: - ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_INT2, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_INT2, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT4 case CCTK_VARIABLE_INT4: - ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_INT4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_INT4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT8 case CCTK_VARIABLE_INT8: - ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_INT8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_INT8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_REAL: - ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_REAL, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_REAL, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_REAL4 case CCTK_VARIABLE_REAL4: - ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_REAL4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_REAL4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL8 case CCTK_VARIABLE_REAL8: - ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_REAL8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_REAL8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL16 case CCTK_VARIABLE_REAL16: - ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_REAL16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_REAL16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_COMPLEX: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -2128,10 +2327,12 @@ int LocalReduce_Norm4_REAL8(int i, int mask_on, CCTK_INT * input_array_offsets, inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_COMPLEX, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_COMPLEX, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -2143,6 +2344,7 @@ int LocalReduce_Norm4_REAL8(int i, int mask_on, CCTK_INT * input_array_offsets, #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -2150,6 +2352,7 @@ int LocalReduce_Norm4_REAL8(int i, int mask_on, CCTK_INT * input_array_offsets, #ifdef CCTK_COMPLEX8 case CCTK_VARIABLE_COMPLEX8: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -2172,10 +2375,12 @@ int LocalReduce_Norm4_REAL8(int i, int mask_on, CCTK_INT * input_array_offsets, inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd8( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd8( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -2187,6 +2392,7 @@ int LocalReduce_Norm4_REAL8(int i, int mask_on, CCTK_INT * input_array_offsets, #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -2195,6 +2401,7 @@ int LocalReduce_Norm4_REAL8(int i, int mask_on, CCTK_INT * input_array_offsets, #ifdef CCTK_COMPLEX16 case CCTK_VARIABLE_COMPLEX16: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -2217,11 +2424,13 @@ int LocalReduce_Norm4_REAL8(int i, int mask_on, CCTK_INT * input_array_offsets, inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -2233,6 +2442,7 @@ int LocalReduce_Norm4_REAL8(int i, int mask_on, CCTK_INT * input_array_offsets, #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -2241,6 +2451,7 @@ int LocalReduce_Norm4_REAL8(int i, int mask_on, CCTK_INT * input_array_offsets, #ifdef CCTK_COMPLEX32 case CCTK_VARIABLE_COMPLEX32: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -2263,11 +2474,13 @@ int LocalReduce_Norm4_REAL8(int i, int mask_on, CCTK_INT * input_array_offsets, inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL8, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -2280,23 +2493,31 @@ int LocalReduce_Norm4_REAL8(int i, int mask_on, CCTK_INT * input_array_offsets, inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) break; #endif } - return num_points; + ierr = Util_TableSetInt(param_table_handle, num_points, "num_points"); + ierr = Util_TableSetReal(param_table_handle, weight_sum, "weight_sum"); + return 0; } #endif #ifdef CCTK_REAL16 -int LocalReduce_Norm4_REAL16(int i, int mask_on, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[]) +int LocalReduce_Norm4_REAL16(int i, int weight_on, const void * const weight, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[], int param_table_handle) { int iter = 0; int sum_indices = 0; - int flag, product, j, k; + int flag, product, j, k, ierr; + + /* Weight variables */ + CCTK_REAL weight_sum = 0.0; + CCTK_REAL weight_value = 1.0; #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -2308,6 +2529,7 @@ int LocalReduce_Norm4_REAL16(int i, int mask_on, CCTK_INT * input_array_offsets, #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -2315,51 +2537,52 @@ int LocalReduce_Norm4_REAL16(int i, int mask_on, CCTK_INT * input_array_offsets, { /* out values type switches*/ case CCTK_VARIABLE_BYTE: - ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_BYTE, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_BYTE, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; case CCTK_VARIABLE_INT: - ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_INT, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_INT, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_INT1 case CCTK_VARIABLE_INT1: - ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_INT1, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_INT1, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT2 case CCTK_VARIABLE_INT2: - ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_INT2, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_INT2, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT4 case CCTK_VARIABLE_INT4: - ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_INT4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_INT4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT8 case CCTK_VARIABLE_INT8: - ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_INT8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_INT8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_REAL: - ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_REAL, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_REAL, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_REAL4 case CCTK_VARIABLE_REAL4: - ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_REAL4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_REAL4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL8 case CCTK_VARIABLE_REAL8: - ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_REAL8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_REAL8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL16 case CCTK_VARIABLE_REAL16: - ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_REAL16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_REAL16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_COMPLEX: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -2382,10 +2605,12 @@ int LocalReduce_Norm4_REAL16(int i, int mask_on, CCTK_INT * input_array_offsets, inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_COMPLEX, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_COMPLEX, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -2397,6 +2622,7 @@ int LocalReduce_Norm4_REAL16(int i, int mask_on, CCTK_INT * input_array_offsets, #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -2404,6 +2630,7 @@ int LocalReduce_Norm4_REAL16(int i, int mask_on, CCTK_INT * input_array_offsets, #ifdef CCTK_COMPLEX8 case CCTK_VARIABLE_COMPLEX8: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -2426,10 +2653,12 @@ int LocalReduce_Norm4_REAL16(int i, int mask_on, CCTK_INT * input_array_offsets, inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd8( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd8( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -2441,6 +2670,7 @@ int LocalReduce_Norm4_REAL16(int i, int mask_on, CCTK_INT * input_array_offsets, #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -2449,6 +2679,7 @@ int LocalReduce_Norm4_REAL16(int i, int mask_on, CCTK_INT * input_array_offsets, #ifdef CCTK_COMPLEX16 case CCTK_VARIABLE_COMPLEX16: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -2471,11 +2702,13 @@ int LocalReduce_Norm4_REAL16(int i, int mask_on, CCTK_INT * input_array_offsets, inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -2487,6 +2720,7 @@ int LocalReduce_Norm4_REAL16(int i, int mask_on, CCTK_INT * input_array_offsets, #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -2495,6 +2729,7 @@ int LocalReduce_Norm4_REAL16(int i, int mask_on, CCTK_INT * input_array_offsets, #ifdef CCTK_COMPLEX32 case CCTK_VARIABLE_COMPLEX32: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -2517,11 +2752,13 @@ int LocalReduce_Norm4_REAL16(int i, int mask_on, CCTK_INT * input_array_offsets, inval.Re = (CCTK_REAL16) typed_vdata[sum_indices]; #endif #define REDUCTION_OPERATION(Norm4, scalar) (scalar).Re = POWER4((scalar).Re); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) (scalar).Re = POWER4((scalar).Re*weight); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_REAL16, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #undef REDUCTION_PREOP_CAST @@ -2533,22 +2770,30 @@ int LocalReduce_Norm4_REAL16(int i, int mask_on, CCTK_INT * input_array_offsets, #define REDUCTION_PREOP_CAST(inval, typed_vdata,sum_indices, out_type) \ inval = (out_type) typed_vdata[sum_indices]; #define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) break; #endif } - return num_points; + ierr = Util_TableSetInt(param_table_handle, num_points, "num_points"); + ierr = Util_TableSetReal(param_table_handle, weight_sum, "weight_sum"); + return 0; } #endif -int LocalReduce_Norm4_COMPLEX(int i, int mask_on, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[]) +int LocalReduce_Norm4_COMPLEX(int i, int weight_on, const void * const weight, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[], int param_table_handle) { int iter = 0; int sum_indices = 0; - int flag, product, j, k; + int flag, product, j, k, ierr; + + /* Weight variables */ + CCTK_REAL weight_sum = 0.0; + CCTK_REAL weight_value = 1.0; #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -2574,7 +2819,8 @@ int LocalReduce_Norm4_COMPLEX(int i, int mask_on, CCTK_INT * input_array_offsets inval = (CCTK_REAL16) POWER2((POWER2((typed_vdata[sum_indices]).Re)+POWER2((typed_vdata[sum_indices]).Im))); #endif -#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + scalar; +#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -2582,51 +2828,52 @@ int LocalReduce_Norm4_COMPLEX(int i, int mask_on, CCTK_INT * input_array_offsets { /* out values type switches*/ case CCTK_VARIABLE_BYTE: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_BYTE, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_BYTE, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; case CCTK_VARIABLE_INT: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_INT, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_INT, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_INT1 case CCTK_VARIABLE_INT1: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_INT1, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_INT1, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT2 case CCTK_VARIABLE_INT2: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_INT2, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_INT2, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT4 case CCTK_VARIABLE_INT4: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_INT4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_INT4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT8 case CCTK_VARIABLE_INT8: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_INT8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_INT8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_REAL: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_REAL, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_REAL, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_REAL4 case CCTK_VARIABLE_REAL4: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_REAL4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_REAL4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL8 case CCTK_VARIABLE_REAL8: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_REAL8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_REAL8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL16 case CCTK_VARIABLE_REAL16: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_REAL16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_REAL16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_COMPLEX: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER2 @@ -2646,10 +2893,12 @@ int LocalReduce_Norm4_COMPLEX(int i, int mask_on, CCTK_INT * input_array_offsets (inval).Re = (CCTK_REAL) (typed_vdata[sum_indices]).Re; (inval).Im = (CCTK_REAL) (typed_vdata[sum_indices]).Im; #define REDUCTION_OPERATION(Norm4, scalar) scalar = CCTK_CmplxMul(scalar, scalar); CCTK_CmplxMul(scalar,scalar); Norm4 = CCTK_CmplxAdd( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) scalar.Re=scalar.Re*weight; scalar.Im=scalar.Im*weight; scalar = CCTK_CmplxMul(scalar, scalar); CCTK_CmplxMul(scalar,scalar); Norm4 = CCTK_CmplxAdd( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_COMPLEX, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_COMPLEX, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -2675,7 +2924,8 @@ int LocalReduce_Norm4_COMPLEX(int i, int mask_on, CCTK_INT * input_array_offsets inval = (CCTK_REAL16) POWER2((POWER2((typed_vdata[sum_indices]).Re)+POWER2((typed_vdata[sum_indices]).Im))); #endif -#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + scalar; +#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -2683,6 +2933,7 @@ int LocalReduce_Norm4_COMPLEX(int i, int mask_on, CCTK_INT * input_array_offsets #ifdef CCTK_COMPLEX8 case CCTK_VARIABLE_COMPLEX8: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -2695,10 +2946,12 @@ int LocalReduce_Norm4_COMPLEX(int i, int mask_on, CCTK_INT * input_array_offsets (inval).Re = (CCTK_REAL4) (typed_vdata[sum_indices]).Re; (inval).Im = (CCTK_REAL4) (typed_vdata[sum_indices]).Im; #define REDUCTION_OPERATION(Norm4, scalar) scalar = CCTK_CmplxMul8(scalar,scalar); scalar = CCTK_CmplxMul8(scalar,scalar);Norm4 = CCTK_CmplxAdd8( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) scalar.Re=scalar.Re*weight; scalar.Im=scalar.Im*weight; scalar = CCTK_CmplxMul8(scalar,scalar); scalar = CCTK_CmplxMul8(scalar,scalar);Norm4 = CCTK_CmplxAdd8( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -2724,7 +2977,8 @@ int LocalReduce_Norm4_COMPLEX(int i, int mask_on, CCTK_INT * input_array_offsets inval = (CCTK_REAL16) POWER2((POWER2((typed_vdata[sum_indices]).Re)+POWER2((typed_vdata[sum_indices]).Im))); #endif -#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + scalar; +#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -2733,6 +2987,7 @@ int LocalReduce_Norm4_COMPLEX(int i, int mask_on, CCTK_INT * input_array_offsets #ifdef CCTK_COMPLEX16 case CCTK_VARIABLE_COMPLEX16: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -2745,11 +3000,13 @@ int LocalReduce_Norm4_COMPLEX(int i, int mask_on, CCTK_INT * input_array_offsets (inval).Re = (CCTK_REAL8) (typed_vdata[sum_indices]).Re; (inval).Im = (CCTK_REAL8) (typed_vdata[sum_indices]).Im; #define REDUCTION_OPERATION(Norm4, scalar) scalar = CCTK_CmplxMul16(scalar,scalar); scalar = CCTK_CmplxMul16(scalar,scalar); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) scalar.Re=scalar.Re*weight; scalar.Im=scalar.Im*weight; scalar = CCTK_CmplxMul16(scalar,scalar); scalar = CCTK_CmplxMul16(scalar,scalar); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -2775,7 +3032,8 @@ int LocalReduce_Norm4_COMPLEX(int i, int mask_on, CCTK_INT * input_array_offsets inval = (CCTK_REAL16) POWER2((POWER2((typed_vdata[sum_indices]).Re)+POWER2((typed_vdata[sum_indices]).Im))); #endif -#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + scalar; +#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -2784,6 +3042,7 @@ int LocalReduce_Norm4_COMPLEX(int i, int mask_on, CCTK_INT * input_array_offsets #ifdef CCTK_COMPLEX32 case CCTK_VARIABLE_COMPLEX32: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -2799,11 +3058,13 @@ int LocalReduce_Norm4_COMPLEX(int i, int mask_on, CCTK_INT * input_array_offsets (inval).Re = (CCTK_REAL16) (typed_vdata[sum_indices]).Re; (inval).Im = (CCTK_REAL16) (typed_vdata[sum_indices]).Im; #define REDUCTION_OPERATION(Norm4, scalar) scalar = CCTK_CmplxMul32(scalar,scalar); scalar = CCTK_CmplxMul32(scalar,scalar); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) scalar.Re=scalar.Re*weight; scalar.Im=scalar.Im*weight; scalar = CCTK_CmplxMul32(scalar,scalar); scalar = CCTK_CmplxMul32(scalar,scalar); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -2829,24 +3090,32 @@ int LocalReduce_Norm4_COMPLEX(int i, int mask_on, CCTK_INT * input_array_offsets inval = (CCTK_REAL16) POWER2((POWER2((typed_vdata[sum_indices]).Re)+POWER2((typed_vdata[sum_indices]).Im))); #endif -#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + scalar; +#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) break; #endif } - return num_points; + ierr = Util_TableSetInt(param_table_handle, num_points, "num_points"); + ierr = Util_TableSetReal(param_table_handle, weight_sum, "weight_sum"); + return 0; } #ifdef CCTK_COMPLEX8 -int LocalReduce_Norm4_COMPLEX8(int i, int mask_on, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[]) +int LocalReduce_Norm4_COMPLEX8(int i, int weight_on, const void * const weight, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[], int param_table_handle) { int iter = 0; int sum_indices = 0; - int flag, product, j, k; + int flag, product, j, k, ierr; + + /* Weight variables */ + CCTK_REAL weight_sum = 0.0; + CCTK_REAL weight_value = 1.0; #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -2872,7 +3141,8 @@ int LocalReduce_Norm4_COMPLEX8(int i, int mask_on, CCTK_INT * input_array_offset inval = (CCTK_REAL16) POWER2((POWER2((typed_vdata[sum_indices]).Re)+POWER2((typed_vdata[sum_indices]).Im))); #endif -#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + scalar; +#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -2880,51 +3150,52 @@ int LocalReduce_Norm4_COMPLEX8(int i, int mask_on, CCTK_INT * input_array_offset { /* out values type switches*/ case CCTK_VARIABLE_BYTE: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_BYTE, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_BYTE, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; case CCTK_VARIABLE_INT: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_INT, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_INT, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_INT1 case CCTK_VARIABLE_INT1: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_INT1, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_INT1, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT2 case CCTK_VARIABLE_INT2: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_INT2, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_INT2, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT4 case CCTK_VARIABLE_INT4: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_INT4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_INT4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT8 case CCTK_VARIABLE_INT8: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_INT8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_INT8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_REAL: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_REAL, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_REAL, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_REAL4 case CCTK_VARIABLE_REAL4: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_REAL4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_REAL4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL8 case CCTK_VARIABLE_REAL8: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_REAL8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_REAL8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL16 case CCTK_VARIABLE_REAL16: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_REAL16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_REAL16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_COMPLEX: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -2937,10 +3208,12 @@ int LocalReduce_Norm4_COMPLEX8(int i, int mask_on, CCTK_INT * input_array_offset (inval).Re = (CCTK_REAL) (typed_vdata[sum_indices]).Re; (inval).Im = (CCTK_REAL) (typed_vdata[sum_indices]).Im; #define REDUCTION_OPERATION(Norm4, scalar) scalar = CCTK_CmplxMul(scalar, scalar); CCTK_CmplxMul(scalar,scalar); Norm4 = CCTK_CmplxAdd( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) scalar.Re=scalar.Re*weight; scalar.Im=scalar.Im*weight; scalar = CCTK_CmplxMul(scalar, scalar); CCTK_CmplxMul(scalar,scalar); Norm4 = CCTK_CmplxAdd( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_COMPLEX, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_COMPLEX, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -2966,7 +3239,8 @@ int LocalReduce_Norm4_COMPLEX8(int i, int mask_on, CCTK_INT * input_array_offset inval = (CCTK_REAL16) POWER2((POWER2((typed_vdata[sum_indices]).Re)+POWER2((typed_vdata[sum_indices]).Im))); #endif -#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + scalar; +#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -2974,6 +3248,7 @@ int LocalReduce_Norm4_COMPLEX8(int i, int mask_on, CCTK_INT * input_array_offset #ifdef CCTK_COMPLEX8 case CCTK_VARIABLE_COMPLEX8: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -2989,10 +3264,12 @@ int LocalReduce_Norm4_COMPLEX8(int i, int mask_on, CCTK_INT * input_array_offset (inval).Re = (CCTK_REAL4) (typed_vdata[sum_indices]).Re; (inval).Im = (CCTK_REAL4) (typed_vdata[sum_indices]).Im; #define REDUCTION_OPERATION(Norm4, scalar) scalar = CCTK_CmplxMul8(scalar,scalar); scalar = CCTK_CmplxMul8(scalar,scalar);Norm4 = CCTK_CmplxAdd8( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) scalar.Re=scalar.Re*weight; scalar.Im=scalar.Im*weight; scalar = CCTK_CmplxMul8(scalar,scalar); scalar = CCTK_CmplxMul8(scalar,scalar);Norm4 = CCTK_CmplxAdd8( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3018,7 +3295,8 @@ int LocalReduce_Norm4_COMPLEX8(int i, int mask_on, CCTK_INT * input_array_offset inval = (CCTK_REAL16) POWER2((POWER2((typed_vdata[sum_indices]).Re)+POWER2((typed_vdata[sum_indices]).Im))); #endif -#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + scalar; +#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -3027,6 +3305,7 @@ int LocalReduce_Norm4_COMPLEX8(int i, int mask_on, CCTK_INT * input_array_offset #ifdef CCTK_COMPLEX16 case CCTK_VARIABLE_COMPLEX16: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3042,11 +3321,13 @@ int LocalReduce_Norm4_COMPLEX8(int i, int mask_on, CCTK_INT * input_array_offset (inval).Re = (CCTK_REAL8) (typed_vdata[sum_indices]).Re; (inval).Im = (CCTK_REAL8) (typed_vdata[sum_indices]).Im; #define REDUCTION_OPERATION(Norm4, scalar) scalar = CCTK_CmplxMul16(scalar,scalar); scalar = CCTK_CmplxMul16(scalar,scalar); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) scalar.Re=scalar.Re*weight; scalar.Im=scalar.Im*weight; scalar = CCTK_CmplxMul16(scalar,scalar); scalar = CCTK_CmplxMul16(scalar,scalar); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3072,7 +3353,8 @@ int LocalReduce_Norm4_COMPLEX8(int i, int mask_on, CCTK_INT * input_array_offset inval = (CCTK_REAL16) POWER2((POWER2((typed_vdata[sum_indices]).Re)+POWER2((typed_vdata[sum_indices]).Im))); #endif -#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + scalar; +#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -3081,6 +3363,7 @@ int LocalReduce_Norm4_COMPLEX8(int i, int mask_on, CCTK_INT * input_array_offset #ifdef CCTK_COMPLEX32 case CCTK_VARIABLE_COMPLEX32: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3093,11 +3376,13 @@ int LocalReduce_Norm4_COMPLEX8(int i, int mask_on, CCTK_INT * input_array_offset (inval).Re = (CCTK_REAL16) (typed_vdata[sum_indices]).Re; (inval).Im = (CCTK_REAL16) (typed_vdata[sum_indices]).Im; #define REDUCTION_OPERATION(Norm4, scalar) scalar = CCTK_CmplxMul32(scalar,scalar); scalar = CCTK_CmplxMul32(scalar,scalar); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) scalar.Re=scalar.Re*weight; scalar.Im=scalar.Im*weight; scalar = CCTK_CmplxMul32(scalar,scalar); scalar = CCTK_CmplxMul32(scalar,scalar); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX8, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3120,24 +3405,32 @@ int LocalReduce_Norm4_COMPLEX8(int i, int mask_on, CCTK_INT * input_array_offset inval = (CCTK_REAL16) POWER2((POWER2((typed_vdata[sum_indices]).Re)+POWER2((typed_vdata[sum_indices]).Im))); #endif -#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + scalar; +#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) break; #endif } - return num_points; + ierr = Util_TableSetInt(param_table_handle, num_points, "num_points"); + ierr = Util_TableSetReal(param_table_handle, weight_sum, "weight_sum"); + return 0; } #endif #ifdef CCTK_COMPLEX16 -int LocalReduce_Norm4_COMPLEX16(int i, int mask_on, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[]) +int LocalReduce_Norm4_COMPLEX16(int i, int weight_on, const void * const weight, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[], int param_table_handle) { int iter = 0; int sum_indices = 0; - int flag, product, j, k; + int flag, product, j, k, ierr; + + /* Weight variables */ + CCTK_REAL weight_sum = 0.0; + CCTK_REAL weight_value = 1.0; #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3163,7 +3456,8 @@ int LocalReduce_Norm4_COMPLEX16(int i, int mask_on, CCTK_INT * input_array_offse inval = (CCTK_REAL16) POWER2((POWER2((typed_vdata[sum_indices]).Re)+POWER2((typed_vdata[sum_indices]).Im))); #endif -#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + scalar; +#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -3171,51 +3465,52 @@ int LocalReduce_Norm4_COMPLEX16(int i, int mask_on, CCTK_INT * input_array_offse { /* out values type switches*/ case CCTK_VARIABLE_BYTE: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_BYTE, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_BYTE, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; case CCTK_VARIABLE_INT: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_INT, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_INT, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_INT1 case CCTK_VARIABLE_INT1: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_INT1, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_INT1, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT2 case CCTK_VARIABLE_INT2: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_INT2, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_INT2, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT4 case CCTK_VARIABLE_INT4: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_INT4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_INT4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT8 case CCTK_VARIABLE_INT8: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_INT8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_INT8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_REAL: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_REAL, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_REAL, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_REAL4 case CCTK_VARIABLE_REAL4: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_REAL4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_REAL4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL8 case CCTK_VARIABLE_REAL8: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_REAL8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_REAL8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL16 case CCTK_VARIABLE_REAL16: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_REAL16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_REAL16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_COMPLEX: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3228,10 +3523,12 @@ int LocalReduce_Norm4_COMPLEX16(int i, int mask_on, CCTK_INT * input_array_offse (inval).Re = (CCTK_REAL) (typed_vdata[sum_indices]).Re; (inval).Im = (CCTK_REAL) (typed_vdata[sum_indices]).Im; #define REDUCTION_OPERATION(Norm4, scalar) scalar = CCTK_CmplxMul(scalar, scalar); CCTK_CmplxMul(scalar,scalar); Norm4 = CCTK_CmplxAdd( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) scalar.Re=scalar.Re*weight; scalar.Im=scalar.Im*weight; scalar = CCTK_CmplxMul(scalar, scalar); CCTK_CmplxMul(scalar,scalar); Norm4 = CCTK_CmplxAdd( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_COMPLEX, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_COMPLEX, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3257,7 +3554,8 @@ int LocalReduce_Norm4_COMPLEX16(int i, int mask_on, CCTK_INT * input_array_offse inval = (CCTK_REAL16) POWER2((POWER2((typed_vdata[sum_indices]).Re)+POWER2((typed_vdata[sum_indices]).Im))); #endif -#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + scalar; +#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -3265,6 +3563,7 @@ int LocalReduce_Norm4_COMPLEX16(int i, int mask_on, CCTK_INT * input_array_offse #ifdef CCTK_COMPLEX8 case CCTK_VARIABLE_COMPLEX8: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3277,10 +3576,12 @@ int LocalReduce_Norm4_COMPLEX16(int i, int mask_on, CCTK_INT * input_array_offse (inval).Re = (CCTK_REAL4) (typed_vdata[sum_indices]).Re; (inval).Im = (CCTK_REAL4) (typed_vdata[sum_indices]).Im; #define REDUCTION_OPERATION(Norm4, scalar) scalar = CCTK_CmplxMul8(scalar,scalar); scalar = CCTK_CmplxMul8(scalar,scalar);Norm4 = CCTK_CmplxAdd8( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) scalar.Re=scalar.Re*weight; scalar.Im=scalar.Im*weight; scalar = CCTK_CmplxMul8(scalar,scalar); scalar = CCTK_CmplxMul8(scalar,scalar);Norm4 = CCTK_CmplxAdd8( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3306,7 +3607,8 @@ int LocalReduce_Norm4_COMPLEX16(int i, int mask_on, CCTK_INT * input_array_offse inval = (CCTK_REAL16) POWER2((POWER2((typed_vdata[sum_indices]).Re)+POWER2((typed_vdata[sum_indices]).Im))); #endif -#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + scalar; +#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -3315,6 +3617,7 @@ int LocalReduce_Norm4_COMPLEX16(int i, int mask_on, CCTK_INT * input_array_offse #ifdef CCTK_COMPLEX16 case CCTK_VARIABLE_COMPLEX16: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3330,11 +3633,13 @@ int LocalReduce_Norm4_COMPLEX16(int i, int mask_on, CCTK_INT * input_array_offse (inval).Re = (CCTK_REAL8) (typed_vdata[sum_indices]).Re; (inval).Im = (CCTK_REAL8) (typed_vdata[sum_indices]).Im; #define REDUCTION_OPERATION(Norm4, scalar) scalar = CCTK_CmplxMul16(scalar,scalar); scalar = CCTK_CmplxMul16(scalar,scalar); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) scalar.Re=scalar.Re*weight; scalar.Im=scalar.Im*weight; scalar = CCTK_CmplxMul16(scalar,scalar); scalar = CCTK_CmplxMul16(scalar,scalar); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3360,7 +3665,8 @@ int LocalReduce_Norm4_COMPLEX16(int i, int mask_on, CCTK_INT * input_array_offse inval = (CCTK_REAL16) POWER2((POWER2((typed_vdata[sum_indices]).Re)+POWER2((typed_vdata[sum_indices]).Im))); #endif -#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + scalar; +#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -3369,6 +3675,7 @@ int LocalReduce_Norm4_COMPLEX16(int i, int mask_on, CCTK_INT * input_array_offse #ifdef CCTK_COMPLEX32 case CCTK_VARIABLE_COMPLEX32: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3381,11 +3688,13 @@ int LocalReduce_Norm4_COMPLEX16(int i, int mask_on, CCTK_INT * input_array_offse (inval).Re = (CCTK_REAL16) (typed_vdata[sum_indices]).Re; (inval).Im = (CCTK_REAL16) (typed_vdata[sum_indices]).Im; #define REDUCTION_OPERATION(Norm4, scalar) scalar = CCTK_CmplxMul32(scalar,scalar); scalar = CCTK_CmplxMul32(scalar,scalar); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) scalar.Re=scalar.Re*weight; scalar.Im=scalar.Im*weight; scalar = CCTK_CmplxMul32(scalar,scalar); scalar = CCTK_CmplxMul32(scalar,scalar); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX16, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3411,24 +3720,32 @@ int LocalReduce_Norm4_COMPLEX16(int i, int mask_on, CCTK_INT * input_array_offse inval = (CCTK_REAL16) POWER2((POWER2((typed_vdata[sum_indices]).Re)+POWER2((typed_vdata[sum_indices]).Im))); #endif -#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + scalar; +#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) break; #endif } - return num_points; + ierr = Util_TableSetInt(param_table_handle, num_points, "num_points"); + ierr = Util_TableSetReal(param_table_handle, weight_sum, "weight_sum"); + return 0; } #endif #ifdef CCTK_COMPLEX32 -int LocalReduce_Norm4_COMPLEX32(int i, int mask_on, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[]) +int LocalReduce_Norm4_COMPLEX32(int i, int weight_on, const void * const weight, CCTK_INT * input_array_offsets, int * indices, int max_iter, int * actual_indices, CCTK_INT * input_array_strides, CCTK_INT * input_array_min_subscripts,const CCTK_INT * input_array_dims, int num_points, int * actual_iters_per_dim, int * iters_per_dim, int N_dims, const void *const input_arrays[], const CCTK_INT output_number_type_codes[], void * const output_numbers[], int param_table_handle) { int iter = 0; int sum_indices = 0; - int flag, product, j, k; + int flag, product, j, k, ierr; + + /* Weight variables */ + CCTK_REAL weight_sum = 0.0; + CCTK_REAL weight_value = 1.0; #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3454,7 +3771,8 @@ int LocalReduce_Norm4_COMPLEX32(int i, int mask_on, CCTK_INT * input_array_offse inval = (CCTK_REAL16) POWER2((POWER2((typed_vdata[sum_indices]).Re)+POWER2((typed_vdata[sum_indices]).Im))); #endif -#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + scalar; +#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -3462,51 +3780,52 @@ int LocalReduce_Norm4_COMPLEX32(int i, int mask_on, CCTK_INT * input_array_offse { /* out values type switches*/ case CCTK_VARIABLE_BYTE: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; case CCTK_VARIABLE_INT: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_INT, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_INT, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_INT1 case CCTK_VARIABLE_INT1: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_INT1, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_INT1, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT2 case CCTK_VARIABLE_INT2: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_INT2, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_INT2, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT4 case CCTK_VARIABLE_INT4: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_INT4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_INT4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_INT8 case CCTK_VARIABLE_INT8: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_INT8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_INT8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_REAL: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_REAL, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_REAL, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #ifdef CCTK_REAL4 case CCTK_VARIABLE_REAL4: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_REAL4, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_REAL4, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL8 case CCTK_VARIABLE_REAL8: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_REAL8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_REAL8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif #ifdef CCTK_REAL16 case CCTK_VARIABLE_REAL16: - ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_REAL16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_REAL16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) break; #endif case CCTK_VARIABLE_COMPLEX: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3519,10 +3838,12 @@ int LocalReduce_Norm4_COMPLEX32(int i, int mask_on, CCTK_INT * input_array_offse (inval).Re = (CCTK_REAL) (typed_vdata[sum_indices]).Re; (inval).Im = (CCTK_REAL) (typed_vdata[sum_indices]).Im; #define REDUCTION_OPERATION(Norm4, scalar) scalar = CCTK_CmplxMul(scalar, scalar); CCTK_CmplxMul(scalar,scalar); Norm4 = CCTK_CmplxAdd( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) scalar.Re=scalar.Re*weight; scalar.Im=scalar.Im*weight; scalar = CCTK_CmplxMul(scalar, scalar); CCTK_CmplxMul(scalar,scalar); Norm4 = CCTK_CmplxAdd( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_COMPLEX, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_COMPLEX, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3548,7 +3869,8 @@ int LocalReduce_Norm4_COMPLEX32(int i, int mask_on, CCTK_INT * input_array_offse inval = (CCTK_REAL16) POWER2((POWER2((typed_vdata[sum_indices]).Re)+POWER2((typed_vdata[sum_indices]).Im))); #endif -#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + scalar; +#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -3556,6 +3878,7 @@ int LocalReduce_Norm4_COMPLEX32(int i, int mask_on, CCTK_INT * input_array_offse #ifdef CCTK_COMPLEX8 case CCTK_VARIABLE_COMPLEX8: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3568,10 +3891,12 @@ int LocalReduce_Norm4_COMPLEX32(int i, int mask_on, CCTK_INT * input_array_offse (inval).Re = (CCTK_REAL4) (typed_vdata[sum_indices]).Re; (inval).Im = (CCTK_REAL4) (typed_vdata[sum_indices]).Im; #define REDUCTION_OPERATION(Norm4, scalar) scalar = CCTK_CmplxMul8(scalar,scalar); scalar = CCTK_CmplxMul8(scalar,scalar);Norm4 = CCTK_CmplxAdd8( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) scalar.Re=scalar.Re*weight; scalar.Im=scalar.Im*weight; scalar = CCTK_CmplxMul8(scalar,scalar); scalar = CCTK_CmplxMul8(scalar,scalar);Norm4 = CCTK_CmplxAdd8( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_COMPLEX8, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3597,7 +3922,8 @@ int LocalReduce_Norm4_COMPLEX32(int i, int mask_on, CCTK_INT * input_array_offse inval = (CCTK_REAL16) POWER2((POWER2((typed_vdata[sum_indices]).Re)+POWER2((typed_vdata[sum_indices]).Im))); #endif -#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + scalar; +#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -3606,6 +3932,7 @@ int LocalReduce_Norm4_COMPLEX32(int i, int mask_on, CCTK_INT * input_array_offse #ifdef CCTK_COMPLEX16 case CCTK_VARIABLE_COMPLEX16: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3618,11 +3945,13 @@ int LocalReduce_Norm4_COMPLEX32(int i, int mask_on, CCTK_INT * input_array_offse (inval).Re = (CCTK_REAL8) (typed_vdata[sum_indices]).Re; (inval).Im = (CCTK_REAL8) (typed_vdata[sum_indices]).Im; #define REDUCTION_OPERATION(Norm4, scalar) scalar = CCTK_CmplxMul16(scalar,scalar); scalar = CCTK_CmplxMul16(scalar,scalar); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) scalar.Re=scalar.Re*weight; scalar.Im=scalar.Im*weight; scalar = CCTK_CmplxMul16(scalar,scalar); scalar = CCTK_CmplxMul16(scalar,scalar); Norm4 = CCTK_CmplxAdd16( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_COMPLEX16, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3648,7 +3977,8 @@ int LocalReduce_Norm4_COMPLEX32(int i, int mask_on, CCTK_INT * input_array_offse inval = (CCTK_REAL16) POWER2((POWER2((typed_vdata[sum_indices]).Re)+POWER2((typed_vdata[sum_indices]).Im))); #endif -#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + scalar; +#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) @@ -3657,6 +3987,7 @@ int LocalReduce_Norm4_COMPLEX32(int i, int mask_on, CCTK_INT * input_array_offse #ifdef CCTK_COMPLEX32 case CCTK_VARIABLE_COMPLEX32: #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3669,11 +4000,13 @@ int LocalReduce_Norm4_COMPLEX32(int i, int mask_on, CCTK_INT * input_array_offse (inval).Re = (CCTK_REAL16) (typed_vdata[sum_indices]).Re; (inval).Im = (CCTK_REAL16) (typed_vdata[sum_indices]).Im; #define REDUCTION_OPERATION(Norm4, scalar) scalar = CCTK_CmplxMul32(scalar,scalar); scalar = CCTK_CmplxMul32(scalar,scalar); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) scalar.Re=scalar.Re*weight; scalar.Im=scalar.Im*weight; scalar = CCTK_CmplxMul32(scalar,scalar); scalar = CCTK_CmplxMul32(scalar,scalar); Norm4 = CCTK_CmplxAdd32( Norm4, scalar); #define REDUCTION_INITIAL(num) (num).Re = 0.0; (num).Im = 0.0; #define EXTRA_STEP(a, b) - ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], mask_on, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) + ITERATE_ON_ARRAY(i,CCTK_COMPLEX32, input_arrays[i], CCTK_COMPLEX32, output_numbers[i], weight_on, weight, input_array_offsets[i], indices, sum_indices, max_iter, iter, flag, actual_indices,input_array_strides, input_array_min_subscripts,input_array_dims,product) #undef REDUCTION_OPERATION +#undef WEIGHTED_REDUCTION_OPERATION #undef REDUCTION_INITIAL #undef EXTRA_STEP #ifdef POWER4 @@ -3699,13 +4032,16 @@ int LocalReduce_Norm4_COMPLEX32(int i, int mask_on, CCTK_INT * input_array_offse inval = (CCTK_REAL16) POWER2((POWER2((typed_vdata[sum_indices]).Re)+POWER2((typed_vdata[sum_indices]).Im))); #endif -#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + scalar; +#define REDUCTION_OPERATION(Norm4, scalar) Norm4 = Norm4 + POWER4(scalar); +#define WEIGHTED_REDUCTION_OPERATION(Norm4, scalar, weight) Norm4 = Norm4 + POWER4(scalar*weight); #define REDUCTION_INITIAL(num) num = 0; #define EXTRA_STEP(a, b) break; #endif } - return num_points; + ierr = Util_TableSetInt(param_table_handle, num_points, "num_points"); + ierr = Util_TableSetReal(param_table_handle, weight_sum, "weight_sum"); + return 0; } #endif |