Simple fix of the current problem - this works only for non-spin-polarized case

Author: alexvakimov

src/libra_py/workflows/nbra/step3.py

@@ -2721,6 +2721,10 @@ def compute_sd_overlaps_in_parallel(step, params):
     active_space2 = ( np.array(active_space) + 1 ).tolist() # conversion to a different convention
     sd_states_reindexed_sorted = params['sd_states_reindexed_sorted']
 
+    #print("active_space = ", active_space)
+    #print("active_space2 = ", active_space2)
+    #print("sd_states_reindexed_sorted = ", sd_states_reindexed_sorted)
+
     t2 = time.time()
     print('Computing the SD overlaps for step', step)
 
@@ -2736,29 +2740,35 @@ def compute_sd_overlaps_in_parallel(step, params):
     # st_ks = data_conv.nparray2MATRIX(st_ks)
     # s_ks_1 = data_conv.nparray2MATRIX(s_ks_1)
     # s_ks_2 = data_conv.nparray2MATRIX(s_ks_2)
+    x = sp.load_npz(F'{res_dir_1}/St_ks_{step+start_time}.npz').todense() 
+    print("original dimensions of the S and ST matrices", x.shape )
+    print("the dimensions of the reduced S and ST matrices", st_ks.shape)
 
     # Computing the overlaps for SDs
     t2 = time.time()
+    #print("sd_states_reindexed_sorted[step] = ", sd_states_reindexed_sorted[step])
+    #print("sd_states_reindexed_sorted[step+1] = ", sd_states_reindexed_sorted[step+1])
+    
     if params['apply_orthonormalization']:
         s_sd_1 = mapping3.ovlp_mat_arb(
             sd_states_reindexed_sorted[step],
             sd_states_reindexed_sorted[step],
             s_ks_1,
-            active_space2)
+            None)
             #use_minimal=False,
             #use_mo_approach=False).real
         s_sd_2 = mapping3.ovlp_mat_arb(
             sd_states_reindexed_sorted[step + 1], 
             sd_states_reindexed_sorted[step + 1], 
             s_ks_2, 
-            active_space2)
+            None)
             #use_minimal=False, 
             #use_mo_approach=False).real
 
     st_sd = mapping3.ovlp_mat_arb(sd_states_reindexed_sorted[step],
                                  sd_states_reindexed_sorted[step + 1],
                                  st_ks,
-                                 active_space2)
+                                 None)
                                  #use_minimal=False,
                                  #use_mo_approach=False).real
     # s_sd_1 = data_conv.MATRIX2nparray(s_sd_1)

src/libra_py/workflows/nbra/step3_many_body.py

@@ -417,7 +417,7 @@ def make_T_matrices_fast(ci_coefficients, ci_basis_states, spin_components, sd_b
     if verbosity >= 1:
         print("=============== In make_T_matrices ============")
         print(F"number_of_states = {number_of_states} nCIs = {nCIs} nSDs = {nSDs}")
-        print(F"sd_basis_states_unique ", sd_bais_states_unique)
+        print(F"sd_basis_states_unique ", sd_basis_states_unique)
 
     # Results
     SD2CI = []