algoskynet / macrodatabase Goto Github PK

It's one thing to understand companies existing in countries, sectors and industries and another to understand movements within these categories due to macroeconomics. The former is attempted to be solved with FinanceDatabase and the MacroDatabase is an attempt to cover the latter.

Within this database of over 570.000 datasets, you are able to view an incredibly large amount of financial indicators (e.g. GDP, employment rates, interest & inflation rates), prices (e.g. commodities, CPIs and housing indices), an extensive variety of country data (e.g. government bond yields of the Netherlands, real consumption of households in Japan) and so much more.

Find an overview of the largest categories within the database here. The categorisation comes from the Federal Reserve Economic Data (FRED) database and this repository aims to make sense of all the data within these categories.

1. Using the Database
2. Questions & Answers
3. Contribution

To access the database you can download the entire repository, but I strongly recommend making use of the package closely attached to the database. It allows you to select specific json files as well as search through collected data with a specific query.

You can install the package with the following steps:

1. pip install macrodatabase
2. (within Python) import macrodatabase as md

The package has the following functions:

• show_country_options(search=None, category=None, parameter=None) - gives curated country options. If no parameter is provided, gives all options. If search is provided, searches with this query and if category or parameter is provided, gives a specific selection.
• show_states_options(search=None, category=None, parameter=None) - gives curated states options. If no parameter is provided, gives all options. If search is provided, searches with this query and if category or parameter is provided, gives a specific selection.
• show_database_options(search=None, include_discontinued=False) - gives all options within the database given a certain search term or all if no search term is provided.
• select_country_data(category, parameter, country=None, unit=None, period=None, seasonality=True) - gives a certain set of curated country data based on the parameters given. E.g. "GDP" and "Liquid Liabilities to GDP". You can also specify the unit, period and seasonality to get specific types of data. If not provided, the function returns all types.
• select_states_data(category, parameter, state=None, unit=None, period=None, seasonality=True) - gives a certain set of curated state data based on the parameters given. E.g. "Real Estate" and "Housing Inventory: Average Listing Price". You can also specify the unit, period and seasonality to get specific types of data. If not provided, the function returns all types.
• select_database_data(parameter) - grab data from the whole database. This follows the structure that can be obtained via show_database_options. E.g. to select Mortgage Rates you would use 'Interest Rates/Mortgage Rates'.
• collect_data(ids, api_key, show_progress=True) - based on IDs, collect data via the FREDAPI package. This allows you to instantly collect all data for the curated list. For example, all countries you can find within 'Liquid Liabilities to GDP' (150+).

In this section you can find answers to commonly asked questions. In case the answer to your question is not here, consider creating an Issue.

• How did you get your data?
  • Please check the Methodology.
• Which countries and states are supported?
  • Please see Countries and States.
• How can I find out which options exists?
  • For this you can use the show_country_options function for the Curated Countries, show_state_options for the Curated States and show_database_options function for general database querying.
• How frequently does the Database get updated?
  • Every sunday this database gets refreshed via GitHub Actions. See Methodology for how this is done.

Projects are bound to have (small) errors and can always be improved. Therefore, we highly encourage you to submit issues and create pull requests to improve the package.