pydeflate 1.0.1

Python package to convert current prices figures to constant prices and vice versa

The pydeflate Package

Pydeflate is a Python package to convert flows data to constant prices. This can be done from any source currency to any desired base year and currency. Pydeflate can also be used to convert constant data to current prices and to convert from one currency to another (in current and constant prices). Users can choose the source of the exchange and deflator/prices data (IMF, World Bank or OECD DAC).

• Free software: MIT license

• Documentation: https://pydeflate.readthedocs.io.

Installation

pydeflate is registered at PyPI. From the command line:

pip install pydeflate --upgrade

Alternatively, the source code is available on GitHub.

Usage

Basic usage

Convert data expressed in current USD prices to constant EUR prices for a given base year:

import pydeflate
import pandas as pd

#example data
data = {'iso_code': ['FRA','USA', 'GTM'],
        'year': [2017, 2017, 2017],
        'value': [50, 100, 200]}

#create an example dataframe, in current USD prices
df = pd.DataFrame.from_dict(data)

#convert to EUR 2015 constant prices
df_constant = pydeflate.deflate(
    df = df,
    base_year = 2015,
    source = 'wb', #exchange/deflators from the IMF
    method = 'gdp',
    source_currency = "USA", #since data is in USD
    target_currency = "EMU", #we want the result in constant EUR
    iso_column = "iso_code",
    date_column = "year",
    source_col = "value",
    target_col = "value_constant",
)

print(df_constant)

This results in a dataframe containing a new column value_constant in 2015 constant prices. In the background, pydeflate takes into account:

• changes in princes, through a gdp deflator in this case

• changes in exchange rates overtime

Pydeflate can also handle data that is expressed in local currency units. In that case, users can specify LCU as the source currency.

import pydeflate
import pandas as pd

#example data
data = {'iso_code': ['GBR','GBR', 'JPN'],
        'date': [2011, 2015, 2015],
        'value': [100, 100, 100]}

#create an example dataframe, in current local currency units
df = pd.DataFrame.from_dict(data)

#convert to USD 2018 constant prices
df_constant = pydeflate.deflate(
    df = df,
    base_year = 2018,
    source = 'imf', #exchange/deflators from the World Bank
    method = 'pcpi',
    source_currency = "LCU", #since data is in LCU
    target_currency = "USA", #to get data in USD
    iso_column = "iso_code",
    date_column = "date",
    source_col = "value",
    target_col = "value", #to not create a new column
)

print(df_constant)

Users can also convert a dataset expressed in constant prices to current prices using pydeflate. To avoid introducing errors, users should know which methodology/ data was used to create constant prices by the original source. The basic usage is the same as before, but the reverse parameter is set to True.

For example, to convert DAC data expressed in 2016 USD constant prices to current US dollars:

import pydeflate
import pandas as pd

#example data
data = {'iso_code': ['USA','ITA', 'ITA'],
        'date': [2010, 2016, 2018],
        'value': [100, 100, 100]}

#create an example dataframe, in current local currency units
df = pd.DataFrame.from_dict(data)

#convert to USD 2018 constant prices
df_current = pydeflate.deflate(
    df = df,
    base_year = 2016,
    source = 'oecd_dac',
    source_currency = "USA", #since data is in USD constant
    target_currency = "USA", #to get data in USD
    iso_column = "iso_code",
    date_column = "date",
    source_col = "value",
    target_col = "value_current",
    reverse = True,
)

print(df_current)

Data source and method options

A source and a method for the exchange and price/gdp deflators must be chosen. The appropriate combination depends on the objectives of the project or the nature of the original data.

In terms of price or GDP deflators, pydeflate provides the following methods:

• World Bank (“wb”):

  • gdp: in order to use GDP deflators.

  • gdp_linked: to use the World Bank’s GDP deflator series which has been linked to produce a consistent time series to counteract breaks in series over time due to changes in base years, sources or methodologies.

  • ‘cpi’: to use Consumer Price Index data

• International Monetary Fund World Economic Outlook (“imf”):

  • pcpi: in order to use Consumer Price Index data.

  • pcpie: to use end-of-period Consumer Price Index data (e.g for December each year).

• OECD Development Assistance Committee (“oecd_dac”):

  • None: for consistency with how the DAC calculates deflators, only their methodology is accepted/used with this data.

The source of the exchange rate data depends on the source selected. Both “imf” and “wb” use data from the International Monetary Fund (LCU per US$, yearly average). The OECD Development Assistance Committee data uses different exchange rates. When oecd_dac is selected as the source, the OECD DAC exchange rates (LCU per US$) are used. Exchange rates between two non USD currency pairs are derived from the LCU to USD exchange rates selected.

Additional features

Pypdeflate relies on data from the World Bank, IMF and OECD for its calculations. This data is updated periodically. If the version of the data stored in the user’s computer is older than 50 days, pydeflate will show a warning on import.

Users can always update the underlying data by using:

import pydeflate

pydeflate.update_all_data()

Pydeflate also provides users with a tool to exchange figures from one currency to another, without applying any deflators. This should only be used on numbers expressed in current prices, however.

In this version of pydeflate, the dataframe must contain a column with iso3 country codes called iso_code.

For example, to convert numbers in current Local Currency Units (LCU) to current Canadian Dollars:

import pydeflate
import pandas as pd

#example data
data = {'iso_code': ['GBR','CAN', 'JPN'],
        'date': [2011, 2015, 2015],
        'value': [100, 100, 100]}

#create an example dataframe, in current local currency units
df = pd.DataFrame.from_dict(data)

#convert to USD 2018 constant prices
df_can = pydeflate.exchange(
    df = df,
    source_currency = "LCU", #since data is in LCU
    target_currency = "CAN", #to get data in Canadian Dollars
    rates_source = 'wb', #this is the same as IMF exchange rates
    value_column = 'value',
    target_column = 'value_CAN',
    date_column = "date",
)

print(df_can)

Credits

This package relies on data from the following sources:

• OECD DAC: https://www.oecd.org/dac/ (Official Development assistance data (DAC1), DAC deflators, and exchange rates used by the DAC)

• IMF World Economic Outlook: https://www.imf.org/en/Publications/WEO (GDP and price deflators)

• World Bank DataBank: https://databank.worldbank.org/home.aspx (exchange rates, GDP and price deflators)

This data is provided based on the terms and conditions set by the orignal sources

This tool was packed for pypi with the help of Cookiecutter and the audreyr/cookiecutter-pypackage project template.

History

1.0.0 (2021-11-27)

• Major release.

This is the first major release of pydeflate.

• This new version effectively breaks any compatibility with previous versions of pydeflate.

• This version is a complete rewrite of the package. Please refer to the documentation for information on how pydeflate works

• The basic functionality of pydeflate can now be considered to be settled. Further releases to pydeflate will extend what is possible, without altering the basic way in which pydeflate works.

0.1.4 (2021-04-21)

• Minor release.

This is a minor update to fix a couple of small errors in doc strings. It also adds unit testing for updating the underlying data.

0.1.3 (2021-04-21)

• Minor release.

This version achieves the basic task at hand. It does not yet have full testing.

0.1.2 (2021-04-21)

• Minor release.

This version achieves the basic task at hand. It does not yet have full testing.

0.1.1 (2021-04-21)

• Minor release.

This version has been yanked.

0.1.0 (2021-04-21)

• First release on PyPI.

This version has been yanked.