When the user calls the column_list_by_type property of a DataFrameWithInfo instance, the columns of the DataFrame are analyzed and split into different sets according to types likenumerical_cols and bool_cols. These sets are partly reorganized and returned into a ColumnListByType instance which contains sets of column names, split into different categories based on types of the values contained (like numerical_cols, bool_cols, med_exam_col_list,...).
The two ColumnListByType attributes numerical_cols and med_exam_col_list are overlapping because inside column_list_by_type property, they are both created as the union of the sets numerical_cols and bool_cols, which are the columns only containing numerical and boolean values respectively.
The suggestion is to modify column_list_by_type property and return an object ColumnListByType with a numerical_cols attribute that does not include bool_cols set and to leave bool_cols columns included inside the med_exam_col_list attribute (which is meant to include all the columns that contain values suitable for numerical analysis (i.e. numerical or boolean values)), and inside bool_cols attribute.
Note: It is important to adapt this and the other repos accordingly (like smvet), especially where numerical_cols attribute is employed.
When the user calls the
column_list_by_typeproperty of aDataFrameWithInfoinstance, the columns of the DataFrame are analyzed and split into different sets according to types likenumerical_colsandbool_cols. These sets are partly reorganized and returned into aColumnListByTypeinstance which contains sets of column names, split into different categories based on types of the values contained (likenumerical_cols,bool_cols,med_exam_col_list,...).The two ColumnListByType attributes
numerical_colsandmed_exam_col_listare overlapping because insidecolumn_list_by_typeproperty, they are both created as the union of the setsnumerical_colsandbool_cols, which are the columns only containing numerical and boolean values respectively.The suggestion is to modify
column_list_by_typeproperty and return an objectColumnListByTypewith anumerical_colsattribute that does not includebool_colsset and to leavebool_colscolumns included inside themed_exam_col_listattribute (which is meant to include all the columns that contain values suitable for numerical analysis (i.e. numerical or boolean values)), and insidebool_colsattribute.Note: It is important to adapt this and the other repos accordingly (like smvet), especially where
numerical_colsattribute is employed.