Utilities for fast timeseries analysis
Project description
pyg-timeseries
pandas is great but pyg-timeseries introduces a few improvements.
- pyg is designed so that for dataframes/series without nans, it matches pandas exactly
- consistent treatments of nan's: unlike pandas, pyg ignores nans everywhere in its calculations.
- np.ndarray and pandas dataframes are treated the same and pyg operates on np.arrays seemlessly
- state-management: pyg introduces a framework for returning not just the timeseries, but also the state of the calculation. This can be fed into the next calculation batch, allowing us not to have to 're-run' everything from the distant past.
- performance-wise, pyg is implemented via numba with performance times comparable to pandas
pip install from https://pypi.org/project/pyg-timeseries/
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
pyg-timeseries-0.0.20.tar.gz
(67.6 kB
view hashes)
Built Distribution
Close
Hashes for pyg_timeseries-0.0.20-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 437170a2b8c9ec27ff1d9df427d4219e39fbcf368ebf76f2099550fbe6453731 |
|
MD5 | 4db716fa1ee4352fd6370884d3bbb986 |
|
BLAKE2b-256 | bfc01f989851d3a6d5af47e7a48a87cb266de7b5e1a6cce6ff26ce5a3d4546f2 |