454 reorganize pipeline decision treepy script

scripts/feedback.json

@@ -0,0 +1,28 @@
+{
+    "mw_saxs": 58.15,
+    "mw_model": 59.5,
+    "mw_err": 0.0,
+    "best_model": "multi_state_model_2_1_1.dat",
+    "overall_chi_square": 1.86,
+    "q_ranges": [
+        0.009046,
+        0.1,
+        0.2,
+        0.36925301
+    ],
+    "chi_squares_of_regions": [
+        2.65,
+        1.86,
+        1.44
+    ],
+    "residuals_of_regions": [
+        -0.086,
+        0.003,
+        0.058
+    ],
+    "mw_feedback": "The difference between the model MW (59.5) and the SAXS MW (58.15) is within acceptable error (2.3%).",
+    "overall_chi_square_feedback": "The overall chi-square of this fit is 1.86. Moderate.",
+    "highest_chi_square_feedback": "The chi-square is highest (2.65) in the region where (0.01 < q < 0.1).",
+    "second_highest_chi_square_feedback": "The 2nd highest chi-square (1.86) is in the region where 0.1 < q < 0.2, but this is okay.",
+    "regional_chi_square_feedback": "The overall structure of the PDB model needs improvement. This may come from a sequence that is off, or the presence of oligomerization states in the sample."
+}

scripts/pipeline_decision_tree.py

@@ -131,58 +131,20 @@ def best_chi_square_i(cs_models, multi_state_models):
     Selects the best chi-square from all multi-state files in the input folder.
 
     Returns the index of the best chi-square in cs_models.
-
-    Chooses the first chi-square with less than 20% error compared to the next value.
-    If no such chi-square exists, returns the index of the minimum chi-square.
+    ["1.7", "1.8", "1.9", "2.5"]
+    ["multi_state_model_1_1_1.dat", "multi_state_model_2_1_1.dat]
+    Selects the highest cs that is <= 2. If they are all > 2, select the lowest.
     """
     print_debug("Comparing chi-squares of all multistates")
 
-    # Round and print for debug purposes
-    cs_models_rounded = [round(cs, 2) for cs in cs_models]
-    print_debug(cs_models_rounded)
-
-    csm_err_threshold = 0.2
-
-    # Handle single value case immediately
     if len(cs_models) == 1:
-        return 0
-
-    # Iterate over pairs of chi-square values
-    for _, (cs_current, cs_next) in enumerate(zip(cs_models, cs_models[1:])):
-        csm_err = abs(cs_next - cs_current) / cs_current
-
-        if csm_err <= csm_err_threshold:
-            return _log_and_return_best(cs_current, cs_models, multi_state_models)
-
-    # If no chi-square meets the threshold, return the smallest one
-    best_cs = min(cs_models)
-    return _log_and_return_best(best_cs, cs_models, multi_state_models)
-
-
-def _log_and_return_best(best_cs, cs_models, multi_state_models):
-    """
-    Logs and returns the index of the best chi-square value.
-    """
-    best_index = cs_models.index(best_cs)
-    multi_states_file = multi_state_models[best_index]
-
-    # Extract the multi-state number
-    multi_states_num = _extract_model_number(multi_states_file)
-
-    print_debug(
-        f"The best chi-square is {round(best_cs, 2)} "
-        f"({multi_states_num} multi states)"
-    )
-    return best_index
-
+        best_cs = cs_models[0]
+    elif any (cs < 2 for cs in cs_models):
+        best_cs = max((cs for cs in cs_models if cs < 2))
+    else:
+        best_cs = min(cs_models)
 
-def _extract_model_number(filename):
-    """
-    Extracts the multi-state model number from the filename.
-    Assumes the pattern 'multi_state_model_#'.
-    """
-    start = filename.find("multi_state_model_") + 18
-    return filename[start]
+    return best_cs
 
 
 def calculate_regional_chi_square_values(
@@ -629,7 +591,7 @@ def handle_poor_fit(
         if second_highest_chi_square_flag == "mid_q_err":
             feedback += (
                 " The movement of flexible regions in the model also do not seem to "
-                "improve the fitting."
+                "improve the fit."
             )
         elif second_highest_chi_square_flag == "high_q_err":
             feedback += " There are also background subtraction problems."
@@ -655,7 +617,7 @@ def handle_poor_fit(
         elif second_highest_chi_square_flag == "mid_q_err":
             feedback += (
                 " The movement of flexible regions in the model also do not seem to "
-                "improve the fitting."
+                "improve the fit."
             )
     else:
         feedback = (