There are various libraries created for Python Time Series. Each of them has its own style, contributors and functions. Each library has usually available own training datasets. This code is testing sktime library. Other available libraries are e.g. flint, darts, tsfresh, arrow, orbit, pastas, pyflux, …
sktime - stable, easy-to-use, flexible, unified framework and modular open-source library for time series machine learning tasks. It offers scikit-learn compatible interfaces, train and test splits, time series regressions, time series classifications (univariate and multivariate), transformations, models, forecasting, it has its own datasets, etc. Probabilistic forecasting is defined for point forecast, variance forecast, quantile forecast, interval forecast, distribution forecast, etc. Sktime creators implemented ridge classifiers for Rocket classifier, forest classifiers for DrCIF classifier, HC2 classifier as very accurate classifier combination. There are available many estimators as well.
pycaret - low code machine learning library not only for time series. Particularly for modelling, regressions, classifications, clustering, anomaly detection, etc.
Use sktime to predict and plot the forecasted number of passengers for the next 36 Quarters :
Use sktime to predict and plot the forecasted unemployment rate for the next 9 Quarters :
1. Python code
2. References
https://www.sktime.org/en/stable/examples/01b_forecasting_proba.html
https://jakevdp.github.io/PythonDataScienceHandbook/04.06-customizing-legends.html
https://pycaret.gitbook.io/docs/
https://tsfresh.readthedocs.io/en/latest/#
https://fredrikj.net/python-flint/
https://orbit-ml.readthedocs.io/en/latest/
https://pythonguides.com/matplotlib-fill_between/
https://arrow.readthedocs.io/en/latest/
https://pastas.readthedocs.io/en/latest/
https://pyflux.readthedocs.io/en/latest/
https://pycaret.gitbook.io/docs/get-started/functions/analyze
https://pycaret.org
https://towardsdatascience.com/pycaret-better-machine-learning-with-python-58b202806d1e
https://towardsdatascience.com/time-series-forecasting-with-pycaret-regression-module-237b703a0c63
https://www.sktime.org/en/stable/index.html
https://www.sktime.org/en/stable/api_reference/datasets.html?highlight=datasets
https://pypi.org/project/pmdarima/
https://towardsdatascience.com/why-start-using-sktime-for-forecasting-8d6881c0a518
https://www.analyticsinsight.net/top-10-python-libraries-for-time-series-analysis-in-2022/
コメント