No description, website, or topics provided.
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
timeseries_utils
.gitignore
LICENSE.md
README.rst
setup.cfg
setup.py

README.rst

Collection of convenience functions for generating some artificial fincance data and testing various prediction techniques

For details see https://arthought.com

Todo list

  1. ✔ add new artificial true multivariate data. Extreme Example: One random series that is repeated in a another series, but timeshifted by n steps, This should not work in arima at all
  2. add new artificial true multivariate data. Extreme Example: One random series that is repeated in a another series, but timeshifted by n steps, This should not work in arima at all
  3. ✕ Stateful (will likely not happen)
  4. Test use of package with real data, e.g. air pollution

Reemark : The series_to_supervised function is inspired by this post: https://machinelearningmastery.com/convert-time-series-supervised-learning-problem-python/

Excellent blog, by the way!