Fixing bug in normalization (ORNL#82)

* fix an indexing bug in denormalization * add min/max loading from pkl
RylieWeaver · Dec 20, 2021 · 84e70f4 · 84e70f4
1 parent 08df317
commit 84e70f4
Show file tree

Hide file tree

Showing 3 changed files with 14 additions and 9 deletions.
diff --git a/hydragnn/postprocess/postprocess.py b/hydragnn/postprocess/postprocess.py
@@ -12,11 +12,10 @@
 
 def output_denormalize(y_minmax, true_values, predicted_values):
     for ihead in range(len(y_minmax)):
-        for isample in range(len(predicted_values[0])):
-            for iatom in range(len(predicted_values[ihead][0])):
-                ymin = y_minmax[ihead][0]
-                ymax = y_minmax[ihead][1]
-
+        ymin = y_minmax[ihead][0]
+        ymax = y_minmax[ihead][1]
+        for isample in range(len(predicted_values[ihead])):
+            for iatom in range(len(predicted_values[ihead][isample])):
                 predicted_values[ihead][isample][iatom] = (
                     predicted_values[ihead][isample][iatom] * (ymax - ymin) + ymin
                 )

diff --git a/hydragnn/run_prediction.py b/hydragnn/run_prediction.py
@@ -22,6 +22,7 @@
 from hydragnn.utils.config_utils import (
     update_config_NN_outputs,
     get_model_output_name_config,
+    normalize_output_config,
 )
 from hydragnn.utils.model import calculate_PNA_degree
 from hydragnn.models.create import create_model_config
@@ -62,6 +63,8 @@ def _(config: dict):
     )
     config = update_config_NN_outputs(config, graph_size_variable)
 
+    config = normalize_output_config(config)
+
     config["NeuralNetwork"]["Architecture"]["input_dim"] = len(
         config["NeuralNetwork"]["Variables_of_interest"]["input_node_features"]
     )

diff --git a/hydragnn/utils/config_utils.py b/hydragnn/utils/config_utils.py
@@ -46,11 +46,14 @@ def update_config_NN_outputs(config, graph_size_variable):
 def normalize_output_config(config):
     var_config = config["NeuralNetwork"]["Variables_of_interest"]
     if "denormalize_output" in var_config and var_config["denormalize_output"]:
-        if "total" in config["Dataset"]["path"].keys():
-            dataset_path = f"{os.environ['SERIALIZED_DATA_PATH']}/serialized_dataset/{config['Dataset']['name']}.pkl"
+        ###loading min/max values from input data file. Only one path is needed
+        if list(config["Dataset"]["path"].values())[0].endswith(".pkl"):
+            dataset_path = list(config["Dataset"]["path"].values())[0]
         else:
-            ###used for min/max values loading below
-            dataset_path = f"{os.environ['SERIALIZED_DATA_PATH']}/serialized_dataset/{config['Dataset']['name']}_train.pkl"
+            if "total" in config["Dataset"]["path"].keys():
+                dataset_path = f"{os.environ['SERIALIZED_DATA_PATH']}/serialized_dataset/{config['Dataset']['name']}.pkl"
+            else:
+                dataset_path = f"{os.environ['SERIALIZED_DATA_PATH']}/serialized_dataset/{config['Dataset']['name']}_train.pkl"
         var_config = update_config_minmax(dataset_path, var_config)
     else:
         var_config["denormalize_output"] = False