Add calculate_scaled_cosine_similarity function to latent_experiments.py

Author: amirrr

.gitignore

@@ -158,3 +158,4 @@ cython_debug/
 #  and can be added to the global gitignore or merged into this file.  For a more nuclear
 #  option (not recommended) you can uncomment the following to ignore the entire idea folder.
 #.idea/
+.DS_Store

latent_experiments/src/latent_experiments.py

@@ -0,0 +1,54 @@
+import pandas as pd
+from sklearn.metrics.pairwise import cosine_similarity
+from sklearn.preprocessing import MinMaxScaler, StandardScaler, RobustScaler, Normalizer, MaxAbsScaler, PowerTransformer
+
+def calculate_scaled_cosine_similarity(data, scale_method='minmax'):
+    """
+    Calculate the scaled cosine similarity matrix for the given data.
+
+    Parameters:
+    - data: The input data to calculate the cosine similarity matrix for.
+    - scale_method: The method to scale the data. Default is 'minmax'.
+
+    Returns:
+    - cosine_similarity_df: The cosine similarity matrix, with the maximum similarity index for each row.
+
+    Raises:
+    - ValueError: If the specified scale method is not recognized.
+
+    """
+
+    # Scale the data according to the specified method
+    scalers = {
+        'minmax': MinMaxScaler(),
+        'standard': StandardScaler(),
+        'robust': RobustScaler(),
+        'l2': Normalizer(norm='l2'),
+        'l1': Normalizer(norm='l1'),
+        'maxabs': MaxAbsScaler(),
+        'yeojohnson': PowerTransformer(method='yeo-johnson')
+    }
+
+    scaler = scalers.get(scale_method.lower())
+
+    if not scaler:
+        raise ValueError(f"Scaling method '{scale_method}' is not recognized. "
+                         "Choose from 'minmax', 'standard', 'robust', 'l2', 'l1', 'maxabs', or 'yeojohnson'.")
+
+    normalized_data = scaler.fit_transform(data)
+
+    normalized_df = pd.DataFrame(normalized_data, columns=data.columns)
+
+    cosine_similarity_matrix = cosine_similarity(normalized_df)
+
+    # Set diagonal to 0 from 1
+    np.fill_diagonal(cosine_similarity_matrix, 0)
+
+    # Convert to DataFrame and add max_index column
+    cosine_similarity_df = pd.DataFrame(cosine_similarity_matrix,
+                                        index=normalized_df.index,
+                                        columns=normalized_df.index)
+    cosine_similarity_df['max_index'] = cosine_similarity_df.idxmax(axis=1)
+
+    return cosine_similarity_df
+