Fix confusion_matrix_df and prep_nn_df functions

.gitignore

@@ -19,4 +19,5 @@ Validate_AE.ipynb
 Train_CovMatch.py
 Train_LinearModels.py
 Validate_AE.py
-.ipynb_checkpoints
+.ipynb_checkpoints
+/ignore/

README.md

@@ -28,4 +28,4 @@ Make sure that you keep up with the latest version of mineralML. To upgrade to t
 pip install mineralML --upgrade
 ```
 
-Mac/Linux installation will be straightforward. Windows installations will require the additional setup of WSL
+Mac/Linux installation will be straightforward. Windows installations will require the additional setup of WSL.

Train_Supervised_Variational.py

@@ -76,13 +76,6 @@
 opx_bad = opx[opx.Mineral != opx.Empirical_Mineral]
 cpx_bad = cpx[cpx.Mineral != cpx.Empirical_Mineral]
 
-
-# %% 
-
-
-ss = StandardScaler()
-array_norm = ss.fit_transform(min_df_lim[oxides])
-
 # %% 
 
 # npz = np.load('parametermatrix_neuralnetwork/best_model_data.npz')
@@ -107,6 +100,26 @@
 # mm.pp_matrix(df_valid_cm, cmap = cmap, savefig = 'test', figsize = (11.5, 11.5))
 # plt.show()
 
+
+# %% 
+
+# Step 2: Read in your DataFrame, drop rows with NaN in specific oxide columns, fill NaNs, and filter minerals
+petrelli_df_load = mm.load_df('Petrelli_cpx.csv')
+petrelli_df, petrelli_df_ex = mm.prep_df_nn(petrelli_df_load)
+petrelli_df_pred, petrelli_probability_matrix = mm.predict_class_prob_nn(petrelli_df)
+
+petrelli_bayes_valid_report = classification_report(
+    petrelli_df_pred['Mineral'], petrelli_df_pred['Predict_Mineral'], zero_division=0
+)
+print("Petrelli Validation Report:\n", petrelli_bayes_valid_report)
+
+petrelli_cm = mm.confusion_matrix_df(petrelli_df_pred['Mineral'], petrelli_df_pred['Predict_Mineral'])
+print("Petrelli Confusion Matrix:\n", petrelli_cm)
+
+petrelli_cm[petrelli_cm < len(petrelli_df_pred['Predict_Mineral'])*0.0005] = 0
+mm.pp_matrix(petrelli_cm) # , savefig = 'none') 
+
+
 # %%
 
 # Step 2: Read in your DataFrame, drop rows with NaN in specific oxide columns, fill NaNs, and filter minerals
@@ -123,7 +136,99 @@
 print("LEPR Confusion Matrix:\n", lepr_cm)
 
 lepr_cm[lepr_cm < len(lepr_df_pred['Predict_Mineral'])*0.0005] = 0
-mm.pp_matrix(lepr_cm, savefig = 'none') 
+mm.pp_matrix(lepr_cm) 
+
+# %% 
+
+
+# Step 2: Read in your DataFrame, drop rows with NaN in specific oxide columns, fill NaNs, and filter minerals
+petdb_df_load = mm.load_df('Validation_Data/PetDB_validationdata_Fe.csv')
+petdb_df, petdb_df_ex = mm.prep_df_nn(petdb_df_load)
+petdb_df_pred, petdb_probability_matrix = mm.predict_class_prob_nn(petdb_df)
+
+petdb_bayes_valid_report = classification_report(
+    petdb_df_pred['Mineral'], petdb_df_pred['Predict_Mineral'], zero_division=0
+)
+print("PetDB Validation Report:\n", petdb_bayes_valid_report)
+
+petdb_cm = mm.confusion_matrix_df(petdb_df_pred['Mineral'], petdb_df_pred['Predict_Mineral'])
+print("PetDB Confusion Matrix:\n", petdb_cm)
+
+petdb_cm[petdb_cm < len(petdb_df_pred['Predict_Mineral'])*0.0005] = 0
+mm.pp_matrix(petdb_cm)
+
+# %% 
+
+# %% 
+
+georoc_df_load = mm.load_df('Validation_Data/GEOROC_validationdata_Fe.csv')
+georoc_df_load['Mineral'] = georoc_df_load['Mineral'].replace('(Al)Kalifeldspar', 'KFeldspar')
+georoc_df, georoc_df_ex = mm.prep_df_nn(georoc_df_load)
+
+georoc_df_pred, georoc_probability_matrix = mm.predict_class_prob_nn(georoc_df)
+
+
+georoc_bayes_valid_report = classification_report(
+    georoc_df_pred['Mineral'], georoc_df_pred['Predict_Mineral'], zero_division=0
+)
+print("GEOROC Validation Report:\n", georoc_bayes_valid_report)
+
+georoc_cm = mm.confusion_matrix_df(georoc_df_pred['Mineral'], georoc_df_pred['Predict_Mineral'])
+print("GEOROC Confusion Matrix:\n", georoc_cm)
+
+georoc_cm[georoc_cm < len(georoc_df_pred['Predict_Mineral'])*0.0005] = 0
+mm.pp_matrix(georoc_cm, savefig = None) 
+
+
+# %% 
+
+cascades_df_load = mm.load_df('Validation_Data/Cascades_CpxAmp_NN.csv')
+cascades_df, cascades_df_ex = mm.prep_df_nn(cascades_df_load)
+
+cascades_df_pred, cascades_probability_matrix = mm.predict_class_prob_nn(cascades_df)
+
+
+cascades_bayes_valid_report = classification_report(
+    cascades_df_pred['Mineral'], cascades_df_pred['Predict_Mineral'], zero_division=0
+)
+print("Cascades Validation Report:\n", cascades_bayes_valid_report)
+
+
+# %% 
+
+cascades_df_pred.to_csv('Validation_Data/Cascades_CpxAmp_NN.csv')
+
+
+# %% 
+
+def confusion_matrix_df_test(given_min, pred_min):
+
+    """
+
+    Constructs a confusion matrix as a pandas DataFrame for easy visualization and 
+    analysis. The function first finds the unique classes and maps them to their 
+    corresponding mineral names. Then, it uses these mappings to construct the 
+    confusion matrix, which compares the given and predicted classes.
+
+    Parameters:
+        given_class (array-like): The true class labels.
+        pred_class (array-like): The predicted class labels.
+
+    Returns:
+        cm_df (DataFrame): A DataFrame representing the confusion matrix, with rows 
+                           and columns labeled by the unique mineral names found in 
+                           the given and predicted class arrays.
+
+    """
+
+    cm_matrix = confusion_matrix(given_min, pred_min)
+    unique, valid_mapping  = mm.unique_mapping_nn(pred_min)
+    cm_df = pd.DataFrame(cm_matrix, index=valid_mapping, columns=valid_mapping)
+
+    return cm_df
+
+
+# %% 
 
 
 # %%