v0.3.2🐛Fix tensor sample stacking issue in sliding window func

v0.3.1 was skipped because we received error "This filename has already been used" from PyPI when releasing v0.3.1, please refer to #25 to know more.

What's Changed

• Fix sliding window func by @WenjieDu in #23
• Release v0.3.1 by @WenjieDu in #24
• Fix error "This filename has already been used" from PyPI when releasing v0.3.1 by @WenjieDu in #25

Full Changelog: v0.3...v0.3.2

v0.3 Add random walk

Add the functions to generation and preprocess random walk datasets.

What's Changed

• Add random walk dataset and release v0.3 by @WenjieDu in #19

Full Changelog: v0.2.2...v0.3

v0.2.2 🐛 Bug fix

Fixed the bug that cannot load UCR&UEA datasets with string labels.

What's Changed

• Encode string labels in UCR&UEA datasets into int type by @WenjieDu in #17

Full Changelog: v0.2.1...v0.2.2

v0.2.1

What's Changed

• Add pyproject.toml, unit testing cases, workflows, issue and pr templates by @WenjieDu in #13

Full Changelog: v0.2...v0.2.1

v0.2 Add Solar Alabama

With Solar Alabama added into BenchPOTS v0.2, we now support the preprocessing of 171 datasets.

What's Changed

• Remove functions from PyPOTS and localize them by @WenjieDu in #7
• Update preprocess_pems_traffic() by @WenjieDu in #8
• Update preprocess_pems_traffic() and add localized functions by @WenjieDu in #9
• Add Solar Alabama processing pipeline and request tsdb>=0.5 by @WenjieDu in #10

Full Changelog: v0.1.1...v0.2

v0.1.1

Update BenchPOTS dependencies.

What's Changed

• Update dependencies and release v0.1.1 by @WenjieDu in #6

Full Changelog: v0.1...v0.1.1

v0.1 Support 170 datasets preprocessing

In benchpots v0.1, we enable this benchmarking suite to preprocess a total of 170 open-source public datasets, including

• 163 UCR&UEA datasets;
• 2 PhysioNet Challenge datasets (2012 and 2019);
• 2 air quality datasets (Beijing Multi-site Air Quality and Italy Air);
• 2 electricity datasets (Electricity Load Diagrams, and Electricity Transformer Temperature);
• PeMS traffic dataset;

What's Changed

• feat: add preprocess_physionet2019; by @yyysjz1997 in #1
• Add feature arg in preprocess_physionet2012 to enable feature selection by @WenjieDu in #2
• Update physionet2019 pipeline by @WenjieDu in #3
• Add seq missing and block missing patterns for dataset generation by @WenjieDu in #4
• Update preprocessing funcs by @WenjieDu in #5

New Contributors

• @yyysjz1997 made their first contribution in #1
• @WenjieDu made their first contribution in #2

Full Changelog: v0.0.1...v0.1

v0.0.1

We release this very initial version with the preprocessing pipeline for PhysioNet-2012 dataset.