Convert income and taxes from roubles to USD using official conversion rate table

convert_to_usd.py

#!/usr/bin/env python3

import pandas as pd
import argparse

def main():
    # Parse command line
    parser = argparse.ArgumentParser(formatter_class=argparse.RawTextHelpFormatter,
                                     description="""
Convert RUR income and taxes, sums them and substracts.
Convesion courses can be found here https://www.cbr.ru/eng/currency_base/dynamics/
Download XLS for the required period and then export it as a CSV.""",
                                     epilog="""
Input CSV files are accepted in the following form:

date,sum
30.11.2000,1000000
31.12.2000,2000000
""")
    parser.add_argument("-c", required=True, dest="courses", help="Courses CSV file for the year")
    parser.add_argument("-i", required=True, dest="income", help="Income CSV file with dates and sums in RUR")
    parser.add_argument("-t", required=True, dest="taxes", help="Taxes CSV file with dates and sums in RUR")
    parser.add_argument("-courses_format", choices=["ddmmyy", "mmddyy"], default="mmddyy", required=False,
                        help="Format of dates in courses data file")
    parser.add_argument("-roubles_format", choices=["ddmmyy", "mmddyy"], default="mmddyy", required=False,
                        help="Format of dates in roubles income and taxes data files")
    args = parser.parse_args()

    # Read data
    courses_names = ["nominal", "date", "rate", "currency"]
    courses_types = {
        "nominal": "int64",
        "date": "str",
        "rate": "float64",
        "currency": "str"
    }
    courses = pd.read_csv(args.courses, header=0, names=courses_names, dtype=courses_types,
                          index_col=1, parse_dates=[1], dayfirst=(args.courses_format == "ddmmyy"))
    roubles_names = ["date", "sum"]
    roubles_types = {
        "date": "str",
        "sum": "float64"
    }
    income = pd.read_csv(args.income, header=0, names=roubles_names, dtype=roubles_types,
                         index_col=0, parse_dates=[0], dayfirst=(args.roubles_format == "ddmmyy"))
    taxes = pd.read_csv(args.taxes, header=0, names=roubles_names, dtype=roubles_types,
                        index_col=0, parse_dates=[0], dayfirst=(args.roubles_format == "ddmmyy"))

    # Do conversions
    def convert(source, rates, name):
        # Sum up all operations for one day
        source = source.groupby(['date']).sum()
        # Fill up missing dates looking up conversion rate for previous dates
        rates_filled = rates.reindex(source.index, method='ffill')
        usd = (source['sum'] / rates_filled.loc[source.index]['rate']).round(decimals=2)
        print(name, "in USD is\n", usd)
        return usd

    usd_income = convert(income, courses, "Income")
    usd_taxes = convert(taxes, courses, "Tax")

    # Sum up income, substract taxes, output everything
    total_income = usd_income.sum()
    print("Total income in USD is", total_income)
    total_taxes = usd_taxes.sum()
    print("Total taxes in USD are", total_taxes)
    print("Income minus taxes: ", total_income - total_taxes)
        
if __name__ == '__main__':
    main()