Minor changes

Author: scemama

org/qmckl_jastrow_champ.org

@@ -11280,23 +11280,23 @@ qmckl_compute_jastrow_champ_factor_een_gl_hpc(const qmckl_context context,
         double cn = c_vector_full[a+n*nucl_num];
         if (cn == 0.0) continue;
 
-        const size_t ishift = elec_num*a;
+        const size_t ishift  = elec_num*a;
         const size_t ishift4 = ishift*4;
 
         const double* restrict tmp_c_amkn  = tmp_c_mkn  + ishift;
         const double* restrict tmp_c_amlkn = tmp_c_mlkn + ishift;
-        const double* restrict een_rescaled_n_amnw  = een_rescaled_n_mnw + ishift;
+        const double* restrict een_rescaled_n_amnw  = een_rescaled_n_mnw  + ishift;
         const double* restrict een_rescaled_n_amlnw = een_rescaled_n_mlnw + ishift;
-        const double* restrict dtmp_c_0amknw  = dtmp_c_mknw + ishift4;
+        const double* restrict dtmp_c_0amknw  = dtmp_c_mknw  + ishift4;
         const double* restrict dtmp_c_0amlknw = dtmp_c_mlknw + ishift4;
         const double* restrict een_rescaled_n_gl_0amnw  = een_rescaled_n_gl_mnw  + ishift4;
         const double* restrict een_rescaled_n_gl_0amlnw = een_rescaled_n_gl_mlnw + ishift4;
 
-        const double* restrict dtmp_c_1amknw =  dtmp_c_0amknw  + elec_num;
+        const double* restrict dtmp_c_1amknw  = dtmp_c_0amknw  + elec_num;
         const double* restrict dtmp_c_1amlknw = dtmp_c_0amlknw + elec_num;
-        const double* restrict dtmp_c_2amknw =  dtmp_c_0amknw  + elec_num2;
+        const double* restrict dtmp_c_2amknw  = dtmp_c_0amknw  + elec_num2;
         const double* restrict dtmp_c_2amlknw = dtmp_c_0amlknw + elec_num2;
-        const double* restrict dtmp_c_3amknw =  dtmp_c_0amknw  + elec_num3;
+        const double* restrict dtmp_c_3amknw  = dtmp_c_0amknw  + elec_num3;
         const double* restrict dtmp_c_3amlknw = dtmp_c_0amlknw + elec_num3;
 
         const double* restrict een_rescaled_n_gl_1amnw  = een_rescaled_n_gl_0amnw  + elec_num;
@@ -11305,6 +11305,7 @@ qmckl_compute_jastrow_champ_factor_een_gl_hpc(const qmckl_context context,
         const double* restrict een_rescaled_n_gl_2amlnw = een_rescaled_n_gl_0amlnw + elec_num2;
         const double* restrict een_rescaled_n_gl_3amnw  = een_rescaled_n_gl_0amnw  + elec_num3;
         const double* restrict een_rescaled_n_gl_3amlnw = een_rescaled_n_gl_0amlnw + elec_num3;
+
         double* const restrict factor_een_gl_1nw = factor_een_gl_0nw + elec_num;
         double* const restrict factor_een_gl_2nw = factor_een_gl_0nw + elec_num2;
         double* const restrict factor_een_gl_3nw = factor_een_gl_0nw + elec_num3;
@@ -11571,7 +11572,7 @@ assert(qmckl_jastrow_champ_provided(context));
      | ~dtmp_c~              | ~double[walk_num][0:cord_num-1][0:cord_num][nucl_num][4][elec_num]~ | in     | vector of non-zero coefficients                |
      | ~een_rescaled_n~      | ~double[walk_num][0:cord_num][nucl_num][elec_num]~                  | in     | Electron-nucleus rescaled factor               |
      | ~een_rescaled_n_gl~   | ~double[walk_num][0:cord_num][nucl_num][4][elec_num]~               | in     | Derivative of Electron-nucleus rescaled factor |
-     | ~factor_een_grad~     | ~double[walk_num][4][elec_num]~                                     | out    | Derivative of Electron-nucleus jastrow         |
+     | ~factor_een_grad~     | ~double[walk_num][3][elec_num]~                                     | out    | Derivative of Electron-nucleus jastrow         |
 
 
      #+begin_src f90 :comments org :tangle (eval f) :noweb yes
@@ -11631,10 +11632,10 @@ function qmckl_compute_jastrow_champ_factor_een_grad_doc( &
            do ii = 1, 3
               do j = 1, elec_num
                  factor_een_grad(j,ii,nw) = factor_een_grad(j,ii,nw) + (           &
-                      tmp_c(j,a,m,k,nw)       * een_rescaled_n_gl(j,ii,a,m+l,nw) + &
-                      (dtmp_c(j,ii,a,m,k,nw))   * een_rescaled_n(j,a,m+l,nw)     + &
-                      (dtmp_c(j,ii,a,m+l,k,nw)) * een_rescaled_n(j,a,m  ,nw)     + &
-                      tmp_c(j,a,m+l,k,nw)     * een_rescaled_n_gl(j,ii,a,m,nw)     &
+                      dtmp_c(j,ii,a,m  ,k,nw) * een_rescaled_n   (j,   a,m+l,nw) + &
+                      dtmp_c(j,ii,a,m+l,k,nw) * een_rescaled_n   (j,   a,m  ,nw) + &
+                      tmp_c(j,a,m  ,k,nw)     * een_rescaled_n_gl(j,ii,a,m+l,nw) + &
+                      tmp_c(j,a,m+l,k,nw)     * een_rescaled_n_gl(j,ii,a,m  ,nw)   &
                       ) * cn
               end do
            end do
@@ -11685,7 +11686,7 @@ end function qmckl_compute_jastrow_champ_factor_een_grad_doc
          const double* een_rescaled_n_gl,
          double* const factor_een_grad );
    #+end_src
-   
+
    #+begin_src c :tangle (eval c) :comments org
 qmckl_exit_code
 qmckl_compute_jastrow_champ_factor_een_grad(const qmckl_context context,
@@ -11798,72 +11799,75 @@ qmckl_compute_jastrow_champ_factor_een_grad_hpc(const qmckl_context context,
       const double* restrict tmp_c_mlkn = tmp_c_mkn + len;
       const double* restrict een_rescaled_n_mnw = &(een_rescaled_n[addr1]);
       const double* restrict een_rescaled_n_mlnw = een_rescaled_n_mnw + len;
-      const double* restrict dtmp_c_mknw = &(dtmp_c[addr0 << 2]);
+      const double* restrict dtmp_c_mknw = &(dtmp_c[addr0*4]);
       const double* restrict dtmp_c_mlknw = dtmp_c_mknw + len4;
-      const double* restrict een_rescaled_n_gl_mnw = &(een_rescaled_n_gl[addr1 << 2]);
+      const double* restrict een_rescaled_n_gl_mnw = &(een_rescaled_n_gl[addr1*4]);
       const double* restrict een_rescaled_n_gl_mlnw = een_rescaled_n_gl_mnw + len4;
 
       for (size_t a = 0; a < (size_t) nucl_num; a++) {
         double cn = c_vector_full[a+n*nucl_num];
         if (cn == 0.0) continue;
 
-        const size_t ishift = elec_num*a;
+        const size_t ishift  = elec_num*a;
         const size_t ishift4 = ishift*4;
 
         const double* restrict tmp_c_amlkn = tmp_c_mlkn + ishift;
         const double* restrict tmp_c_amkn = tmp_c_mkn + ishift;
-        const double* restrict een_rescaled_n_amnw = een_rescaled_n_mnw + ishift;
+        const double* restrict een_rescaled_n_amnw  = een_rescaled_n_mnw  + ishift;
         const double* restrict een_rescaled_n_amlnw = een_rescaled_n_mlnw + ishift;
-        const double* restrict dtmp_c_0amknw = dtmp_c_mknw + ishift4;
+        const double* restrict dtmp_c_0amknw  = dtmp_c_mknw  + ishift4;
         const double* restrict dtmp_c_0amlknw = dtmp_c_mlknw + ishift4;
-        const double* restrict een_rescaled_n_gl_0amnw = een_rescaled_n_gl_mnw + ishift4;
+        const double* restrict een_rescaled_n_gl_0amnw  = een_rescaled_n_gl_mnw  + ishift4;
         const double* restrict een_rescaled_n_gl_0amlnw = een_rescaled_n_gl_mlnw + ishift4;
 
-        const double* restrict dtmp_c_1amknw =  dtmp_c_0amknw  + elec_num;
+        const double* restrict dtmp_c_1amknw  = dtmp_c_0amknw  + elec_num;
         const double* restrict dtmp_c_1amlknw = dtmp_c_0amlknw + elec_num;
-        const double* restrict dtmp_c_2amknw =  dtmp_c_0amknw  + elec_num2;
+        const double* restrict dtmp_c_2amknw  = dtmp_c_0amknw  + elec_num2;
         const double* restrict dtmp_c_2amlknw = dtmp_c_0amlknw + elec_num2;
+
         const double* restrict een_rescaled_n_gl_1amnw  = een_rescaled_n_gl_0amnw  + elec_num;
         const double* restrict een_rescaled_n_gl_1amlnw = een_rescaled_n_gl_0amlnw + elec_num;
         const double* restrict een_rescaled_n_gl_2amnw  = een_rescaled_n_gl_0amnw  + elec_num2;
         const double* restrict een_rescaled_n_gl_2amlnw = een_rescaled_n_gl_0amlnw + elec_num2;
+
         double* const restrict factor_een_grad_1nw = factor_een_grad_0nw + elec_num;
         double* const restrict factor_een_grad_2nw = factor_een_grad_0nw + elec_num2;
 
 #ifdef HAVE_OPENMP
 #pragma omp simd
 #endif
         for (size_t j = 0; j < (size_t) elec_num; ++j) {
-          factor_een_grad_0nw[j] = factor_een_grad_0nw[j] + cn *
-            (tmp_c_amkn[j] * een_rescaled_n_gl_0amlnw[j] +
-             dtmp_c_0amknw[j] * een_rescaled_n_amlnw[j]       +
-             dtmp_c_0amlknw[j] * een_rescaled_n_amnw[j]       +
-             tmp_c_amlkn[j] * een_rescaled_n_gl_0amnw[j]);
+          factor_een_grad_0nw[j] = factor_een_grad_0nw[j] + cn * (
+             dtmp_c_0amknw [j] * een_rescaled_n_amlnw[j]  +
+             dtmp_c_0amlknw[j] * een_rescaled_n_amnw [j]  +
+             tmp_c_amkn [j] * een_rescaled_n_gl_0amlnw[j] +
+             tmp_c_amlkn[j] * een_rescaled_n_gl_0amnw [j] );
+
         }
 
 #ifdef HAVE_OPENMP
 #pragma omp simd
 #endif
         for (size_t j = 0; j < (size_t) elec_num; ++j) {
-          factor_een_grad_1nw[j] = factor_een_grad_1nw[j] + cn *
-            (tmp_c_amkn[j] * een_rescaled_n_gl_1amlnw[j] +
-             dtmp_c_1amknw[j] * een_rescaled_n_amlnw[j]       +
-             dtmp_c_1amlknw[j] * een_rescaled_n_amnw[j]       +
-             tmp_c_amlkn[j] * een_rescaled_n_gl_1amnw[j]);
+          factor_een_grad_1nw[j] = factor_een_grad_1nw[j] + cn * (
+             dtmp_c_1amknw [j] * een_rescaled_n_amlnw[j]  +
+             dtmp_c_1amlknw[j] * een_rescaled_n_amnw [j]  +
+             tmp_c_amkn [j] * een_rescaled_n_gl_1amlnw[j] +
+             tmp_c_amlkn[j] * een_rescaled_n_gl_1amnw [j]);
         }
 
 #ifdef HAVE_OPENMP
 #pragma omp simd
 #endif
         for (size_t j = 0; j < (size_t) elec_num; ++j) {
-          factor_een_grad_2nw[j] = factor_een_grad_2nw[j] + cn *
-            (tmp_c_amkn[j] * een_rescaled_n_gl_2amlnw[j] +
-             dtmp_c_2amknw[j] * een_rescaled_n_amlnw[j]       +
-             dtmp_c_2amlknw[j] * een_rescaled_n_amnw[j]       +
-             tmp_c_amlkn[j] * een_rescaled_n_gl_2amnw[j]);
+          factor_een_grad_2nw[j] = factor_een_grad_2nw[j] + cn * (
+             dtmp_c_2amknw [j] * een_rescaled_n_amlnw[j]  +
+             dtmp_c_2amlknw[j] * een_rescaled_n_amnw [j]  +
+             tmp_c_amkn [j] * een_rescaled_n_gl_2amlnw[j] +
+             tmp_c_amlkn[j] * een_rescaled_n_gl_2amnw [j]);
         }
 
-      } 
+      }
     }
   }
   return info;
@@ -12973,7 +12977,8 @@ assert(qmckl_jastrow_champ_provided(context));
       for (int64_t e=0 ; e<elec_num; ++e) {
         if (fabs(total_j_grad[k][m][e] - total_j_deriv[k][m][e]) > 1e-12) {
           printf("%ld %ld %ld\n", k, m, e);
-          printf("total_j_grad = %20.15e\n", total_j_grad[k][m][e]);
+          printf("total_j_deriv = %20.15e\n", total_j_deriv[k][m][e]);
+          printf("total_j_grad  = %20.15e\n", total_j_grad[k][m][e]);
           fflush(stdout);
         }
         assert (fabs(total_j_deriv[k][m][e] - total_j_grad[k][m][e]) < 1.e-12);