Feature/pip support

src/midrc_react/core/jsdcontroller.py

@@ -324,11 +324,8 @@ def build_date_list(df1, df2):
                 str_col = category[:-6]
                 num_col = data_source_1.numeric_cols[str_col]['raw column']
 
-                # remove the 'Not Reported' data before calculation
-                combined_df = combined_df[pd.to_numeric(combined_df[num_col], errors='coerce').notnull()]
-
-                input_data = [float(calc_ks2_samp_by_feature(combined_df[combined_df['date'] <= date],
-                                                             num_col)['Dataset 0 vs Dataset 1']) for date in date_list]
+                input_data = [calc_ks2_samp_by_feature(combined_df[combined_df['date'] <= date],
+                                                             num_col)['Dataset 0 vs Dataset 1'] for date in date_list]
 
             if input_data is not None:
                 model_input_data.append([pandas_date_to_qdate(calc_date) for calc_date in date_list])

src/midrc_react/core/jsdmodel.py

@@ -19,12 +19,19 @@
 
 from typing import Any, Optional
 
+from numpy import float64 as np_float64
 from PySide6.QtCore import QAbstractTableModel, QModelIndex, Qt, Signal
 from PySide6.QtGui import QColor
 
 from midrc_react.core.excel_layout import DataSource
 
 
+def convert_to_builtin(val):
+    if isinstance(val, np_float64):
+        return float(val)
+    return val
+
+
 class JSDTableModel(QAbstractTableModel):
     r"""
     A class representing a table model for JSD data.
@@ -198,7 +205,7 @@ def data(self, index: QModelIndex, role: int = Qt.DisplayRole) -> Optional[Any]:
         """
         if role in (Qt.DisplayRole, Qt.EditRole):
             try:
-                return self._input_data[index.column()][index.row()]
+                return convert_to_builtin(self._input_data[index.column()][index.row()])
             except IndexError:
                 return None
         elif role == Qt.BackgroundRole:

src/midrc_react/core/numeric_distances.py

@@ -29,6 +29,18 @@
 from midrc_react.core.aggregate_jsd_calc import calc_jsd_from_counts_dict
 from midrc_react.core.cucconi import cucconi_test
 
+def remove_nan_from_df(df, column_name):
+    """
+    Remove NaN values from a specific column in a DataFrame.
+
+    Args:
+        df (pd.DataFrame): The DataFrame to process.
+        column_name (str): The name of the column to check for NaN values.
+
+    Returns:
+        pd.DataFrame: A DataFrame with NaN values removed from the specified column.
+    """
+    return df[pd.to_numeric(df[column_name], errors='coerce').notnull()]
 
 def calc_numerical_metric_by_feature(df, feature: str, dataset_column: str, metric_function):
     """
@@ -91,7 +103,7 @@ def calc_cucconi_by_feature(df, feature: str, dataset_column: str = '_dataset_',
     """
     if scaling is not None:
         logging.warning('Cucconi test is not affected by scaling. Ignoring scaling method.')
-    calc_df = df  # if scaling is None else scale_feature(df, feature, method=scaling)
+    calc_df = remove_nan_from_df(df, feature)  # if scaling is None else scale_feature(df, feature, method=scaling)
 
     def cucconi_2samp_test(values1, values2):
         return cucconi_test(values1, values2, method='permutation')
@@ -119,7 +131,7 @@ def calc_ks2_samp_by_feature(df, feature: str, dataset_column: str = '_dataset_'
     """
     if scaling is not None:
         logging.warning('Kolmogorov-Smirnov test is not affected by scaling. Ignoring scaling method.')
-    calc_df = df  # if scaling is None else scale_feature(df, feature, method=scaling)
+    calc_df = remove_nan_from_df(df, feature)  # if scaling is None else scale_feature(df, feature, method=scaling)
     return calc_numerical_metric_by_feature(calc_df, feature, dataset_column, ks_2samp)
 
 
@@ -147,6 +159,7 @@ def calc_wasserstein_by_feature(df, feature: str, dataset_column: str = '_datase
     # We need to define a function that returns a SimpleNamespace with a 'statistic' attribute
     def w_d_calc(values1, values2):
         return SimpleNamespace(statistic=wasserstein_distance(values1, values2))
+    df = remove_nan_from_df(df, feature)
     calc_df = df if scaling is None else scale_feature(df, feature, method=scaling)
     return calc_numerical_metric_by_feature(calc_df, feature, dataset_column, w_d_calc)