utils_df.py

import pandas as pd
import numpy as np
import re
from constants import ColumnNames
import utils


def add_categories_to_df(df, categories_dict):
    for category_name, keywords in categories_dict.items():
        pattern = '|'.join(re.escape(keyword) for keyword in keywords)
        df['category'] = np.where(df['category'].str.strip() == '',
                                  np.where(df[ColumnNames.TEXT].str.contains(pattern, case=False, na=False),
                                           category_name, df['category']), df['category'])


def delete_rows(df, to_del_list):
    pattern = '|'.join(re.escape(item) for item in to_del_list)
    return df[~df[ColumnNames.TEXT].str.contains(pattern, case=False, na=False)]


def get_monthly_expense_df(df, df_grouped):
    df['month'] = utils.get_date_col_as_datetime(df).dt.to_period('M').astype(str)
    monthly_expenses = df.groupby(['month', ColumnNames.CATEGORY])[ColumnNames.AMOUNT].sum().reset_index()
    category_order = df_grouped.sort_values(by=ColumnNames.AMOUNT, ascending=False)[ColumnNames.CATEGORY].tolist()
    monthly_expenses[ColumnNames.CATEGORY] = pd.Categorical(monthly_expenses[ColumnNames.CATEGORY],
                                                            categories=category_order,
                                                            ordered=True)
    return monthly_expenses