pebsconsulting / python_api_client

A Python client for Calcbench's API.

Calcbench Client

A Python client for Calcbench's API.

Calcbench normalizes the data XBRL tagged accounting metrics in the 10-K and 10-Q documents public companies file the with SEC. If you are spending a lot of time on Edgar and know some Python this package might make your life easier.

Get a free two week Calcbench trial @ https://www.calcbench.com/join.

Your Calcbench username (your email) and Calcbench password are your credentials for this package.

See examples @ https://www.calcbench.com/home/api

To install the client with pip use:

pip install git+git://github.com/calcbench/python_api_client.git

To set your credentials either set CALCBENCH_USERNAME and CALCBENCH_PASSWORD environment variables or call:

calcbench.set_credentials({calcbench_username}, {calcbench_password})

To get normalized data call normalized_dataframe, for instance:

calcbench.normalized_dataframe(company_identifiers=['msft', 'ibm'], metrics=['revenue', 'assets'], start_year=2010, start_period=1, end_year=2014, end_period=4)

To get 'As Reported' statements, call as_reported_raw, for instance:

calcbench.as_reported('msft', 'income')

To get breakout/segments call breakouts_raw, for instance:

calcbench.breakouts_raw(company_identifiers=['MSFT', 'AXP'], metrics=['operatingSegmentRevenue', 'operatingSegmentAssets'])

Company identifiers, tickers in most cases, can be retrieved by Standard Industrial Classification (SIC) code or index, for instance

calcbench.tickers(index='DJIA')

Search for footnotes/disclosures, for instance to search for "going concern" in coal company filings:

coal_companies = cb.tickers(SIC_codes=[1200])

cb.text_search(company_identifiers=coal_companies, full_text_search_term='"going concern"', year=2015, period=0)

This is a work in progress. Let me know if you have suggestions or encounter bugs, [email protected].

Credit

https://github.com/calcbench/python_api_client/commit/6c2312525fa365acc91bd8e979037fc2492845f3 https://github.com/someben