Simfin

Public beta version.

Please be patient if you experience any problems. Although the system may look
fully automated, there is actually a little man sitting inside the server to
handle all of your requests: https://youtu.be/8cXF88XzALk

Changes

- Added function `sf.max_drawdown()`

- Added documentation using the sphinx system and setup automatic generation
of the docs at: https://simfin.readthedocs.io/

- Setup automatic daily testing using GitHub Actions for both the simfin
package and the tutorials.

Early public beta version.

In this version, we were mainly trying to decide which colour the icons
should be: https://youtu.be/UMXs9i201AQ

Changes

- Added `sf.StockHub` for easy loading and processing of stock data,
and wrote a new tutorial on this.

- Added function `sf.rel_change()` for calculating stock returns and
various growth-rates. Added `sf.mean_log_change()` for calculating
mean-log stock returns e.g. for all 1-3 year periods. Also wrote a
new tutorial for these functions.

- Added function `sf.apply()` for applying a function to a DataFrame
with either a single or multiple stocks. Changed `sf.asfreq()`,
`sf.resample()`, `sf.reindex()`, and `sf.rel_change()` to use this.
Added a section to Tutorial 01 on this.

- Added several functions for calculating signals from financial data,
and wrote a new tutorial for these functions.

- Added wrapper-function ` sf.cache` for caching the result of slow
functions to disk. Added its use in all signal-functions. Wrote
Tutorial 06 on this.

- Extended bulk-data system to support json-files with meta-info.
Added functions `sf.load_info_datasets` and `sf.load_info_columns`
and used them in `datasets.py` and `test_bulk_data.py`. Also added
functions `sf.info_datasets` and `sf.info_columns` and used them
in Tutorial 01.

"Private" beta version only intended for a small number of friends and
experienced users of SimFin.

It took us a while to make these changes, because we consider our moves
very carefully: https://youtu.be/HxWCAJS7Co8

Changes

- Added bulk-data load-functions for banks and insurance companies.
- Changed all specialized bulk-data load-functions to use functools.partial.
- Added 'full' variants for datasets: income, balance, cashflow.
The 'full' variant has more columns than the 'free' variant.
- Added 'shareprices-latest' dataset variant.
- Added support for dataset markets e.g. `sf.load_income(variant='annual', market='us')`
- Added a script that generates 'names.py' from the SimFin-server's database.
- Improved implementation of `sf.resample()` and also added the functions
`sf.reindex()` and `sf.asfreq()`. Also wrote a new tutorial for these.

Alpha version only intended for developers.