v0.3.1

A bug in the calculation of the delta matrix (time-decay matrix) discussed in #294 gets fixed in this update.

What's Changed

• Update logo URLs by @WenjieDu in #293
• Fixing the issue in delta calculation by @WenjieDu in #297
• Fixing the issue in time-decay matrix calculation and simplify the code by @WenjieDu in #298
• Roll back the delta calculation of M-RNN to the same with GRU-D by @WenjieDu in #300

Full Changelog: v0.3...v0.3.1

v0.3 coming with new features 😎

Happy New Year, dear friends! 🥳

New features and updated APIs in PyPOTS are brought to you here! In v0.3, we

1. added TimesNet as an imputation model;
2. simplified the structure of val_set. In previous versions, you had to give indicating_mask in the dictionary val_set that tells PyPOTS to use which values to validate the model. Now you only need to give X_ori (i.e. X_intact before) and X, both leaving their missing data as NaNs. PyPOTS will handle everything left to evaluate the model for you;
3. enabled PyPOTS to tune hyperparameters for external models (implemented with the PyPOTS framework but haven't been integrated into PyPOTS);
4. updated the package pypots.data.saving. Separated the functions for pickle saving and h5py saving, and added load_dict_from_h5 that can inverse (deserialize) the process of save_dict_into_h5;
5. fixed some bugs (#255, #263, #266, #280, #282, #286, #289);

What's Changed

• Code refactor by @WenjieDu in #251
• Adding TimesNet as an imputation model by @WenjieDu in #252
• Adding TimesNet, refactoring code, and updating docs by @WenjieDu in #253
• Fixing CSDI gtmask bug by @WenjieDu in #255
• Fixing CSDI gt_mask issue, and setting a fixed random seed for testing cases by @WenjieDu in #256
• Making CSDI return all n_sampling_times imputation samples by @WenjieDu in #258
• Adding get_random_seed(), and adding func calc_quantile_crps() by @WenjieDu in #260
• Making CSDI val process same as the original by @WenjieDu in #262
• Fix missing argument attn_dropout in imputation Transformer by @WenjieDu in #263
• Adding visualization functions by @AugustJW in #267
• Add cluster plotting functions in pypots.utils.visualization by @vemuribv in #182
• Fixing unstable nonstationary norm, adding utils.visual, and doing some code refactoring by @WenjieDu in #266
• Updating package pypots.data.saving by @WenjieDu in #268
• Enabling to tune hyperparameters for outside models implemented with PyPOTS framework by @WenjieDu in #269
• Simplifying the structure of val_set, and using a consistent strategy when lazy-loading val_set by @WenjieDu in #272
• Renaming X_intact into X_ori, and adding matplotlib as a dependency by @WenjieDu in #274
• Simplifying val_set, renaming X_intact, and adding unit tests for the visual package by @WenjieDu in #275
• Update GP-VAE by @WenjieDu in #277
• Updating GP-VAE, adding load_dict_from_h5, etc. by @WenjieDu in #278
• Adding _check_inputs() for error calculation functions by @WenjieDu in #279
• Fixing CSDI, adding placeholder for epoch num in logging by @WenjieDu in #280
• Fixing the infinite loop in LOCF by @WenjieDu in #282
• Update docs by @WenjieDu in #285
• Updating docs, fixing CSDI&LOCF&MRNN, and adding the strategy to save all models by @WenjieDu in #284
• Making PyPOTS able to save all models during training, checking if d_model=n_heads*d_k for SAITS and Transformer by @WenjieDu in #287
• Fixing MRNN by @WenjieDu in #286
• Fix issues in MRNN and update the hyperparameter tuning functionality by @WenjieDu in #288
• Fixing the type error of random_seed in pypots.cli.tuning and updating the docs by @WenjieDu in #289
• Updating load_dict_from_h5() by @WenjieDu in #290

Full Changelog: v0.2.1...v0.3

v0.2.1

Here are updates,

1. for missing values after LOCF imputation (that are missing since the first step hence LOCF doesn't work), we added more options to handle them. Please refer to the argument first_step_imputation in LOCF docs. The default option is "zero" in previous versions, but we've changed it to "backward" which is more reasonable;
2. enabled SAITS to return latent attention weights from blocks in predict() for advanced analysis e.g. in #178;
3. renamed model saving and loading functions save_model() and load_model() into save() and load();

What's Changed

• Check if X_intact contains missing data for imputation models, check and list mismatched hyperparameters in the tuning mode by @WenjieDu in #234
• Make SAITS return attention weights in predict() by @WenjieDu in #239
• Adding other options for the first step imputation in LOCF by @WenjieDu in #240
• Fixing the problem about staling issues by @WenjieDu in #244
• Testing with Python 3.11 and support it by @WenjieDu in #246
• Rename save_model() and load_model() into save() and load() by @WenjieDu in #247
• Refactoring save_model() and load_model(), and updating docs by @WenjieDu in #249

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

PyPOTS v0.2 🤗

In PyPOTS v0.2 this new version, we

1. enabled hyperparameter tuning for all NN algorithms;
2. fixed a bug in the updating strategy of term F in CRLI;
3. replaced the license GPL-v3 with BSD-3-Clause that has less constraints;
4. announced PyPOTS ecosystem;

What's Changed

• Adding tsdb and pygrinder into the docs by @WenjieDu in #212
• Fixing disappeared TSDB and PyGrinder by @WenjieDu in #213
• Replacing PyCorruptor with PyGrinder by @WenjieDu in #215
• Merge the docs of PyPOTS ecosystem, and replace pycorruptor with pygrinder in pypots by @WenjieDu in #216
• Clone TSDB and PyGrinder repos to use their latest code and docs by @WenjieDu in #217
• Install from source code to use the latest docs of TSDB and PyGrinder by @WenjieDu in #219
• Install from TSDB and PyGrinder repos to use their latest docs by @WenjieDu in #220
• Enable hyperparameter tuning with NNI framework by @WenjieDu in #221
• Fixing dependency error in testing_CI workflow by @WenjieDu in #223
• Enable hyperparameter tuning with NNI, fix dependency error in testing_CI, and update docs by @WenjieDu in #224
• Fix the bug in the updating strategy of term F in CRLI by @WenjieDu in #226
• Apply BSD-3 license, and update docs by @WenjieDu in #228
• Fix a bug in CRLI, switch to BSD-3 license by @WenjieDu in #229
• Update version limitations on dependencies, and install dependencies in PyPI publishing workflow by @WenjieDu in #230

Full Changelog: v0.1.4...v0.2

v0.1.4

In this new version, we made the following changes:

1. added the imputation model CSDI;
2. added the unified method predict() for all models to run inference on the given test set;
3. enabled clustering algorithms to select the best model on the validation set;
4. fixed the bug that GP-VAE failed to run on CUDA devices;
5. made SAITS to use customized loss function specified by users;

What's Changed

• Add the method predict() for all models by @WenjieDu in #199
• Refactor algorithms' module structure, enable customized loss function in SAITS, enable GP-VAE to run on CUDA, etc. by @WenjieDu in #201
• Merge dev into main by @WenjieDu in #202
• Make clustering algorithms to select the best model according to the loss on a given validation set by @WenjieDu in #204
• Fixing failed CI testing due to dependency installation error by @WenjieDu in #206
• Adding the model CSDI by @WenjieDu in #208
• Refactoring, and updating the docs by @WenjieDu in #209
• Adding CSDI, updating the docs by @WenjieDu in #210

Full Changelog: v0.1.3...v0.1.4

v0.1.3

In this release, we mainly

1. added new models US-GAN and GP-VAE for the imputation task on POTS data;
2. appended util functions save_dict_into_h5() to help users directly save a dataset from a Python Dictionary into an HDF5 file that can be used by models in PyPOTS with lazy-loading strategy to save RAM and CUDA memory;
3. added internal clustering validation measures to help evaluate clustering models;
4. enabled PyPOTS clustering models to return their latent representation for advanced analysis, e.g. internal clustering validation;
5. added learning-rate schedulers for optimizers;

👍 Kudos to our new contributors, @AugustJW and @vemuribv! The above points 1 and 3 are implemented by them that enrich this release and make it happen here.

😎 More awesome features are on the road. Stay tuned or join us to make them happen!

What's Changed

• Adding the package to store frequently-used modules; by @WenjieDu in #170
• feat: add Transformer encoder, decoder, and decoder_layer; by @WenjieDu in #172
• Adding the package pypots/modules, and appending modules of vanilla Transformer by @WenjieDu in #173
• Add internal clustering validation measures to pypots.utils.metrics by @vemuribv in #179
• Add save_dict_into_h5() and sliding_window(), add .gitignore, and update the docs by @WenjieDu in #180
• Merge dev into pr/176 by @WenjieDu in #181
• Fixing failed doc build by @WenjieDu in #186
• Fixing the failed workflow Autoreply to Issues Opened for PyPOTS and update the PR template by @WenjieDu in #184
• Merge dev into main by @WenjieDu in #188
• add models GP-VAE/USGAN by @AugustJW in #176
• Refactor testing cases by @WenjieDu in #189
• Add models US-GAN and GP-VAE, update docs, refactor testing cases, add cal_internal_cluster_validation_metrics() by @WenjieDu in #190
• Extract latent from clustering models by @WenjieDu in #192
• Add internal and external cluster validation funcs, and enable CRLI and VaDER to return latent for advanced analysis by @WenjieDu in #193
• Add learning-rate schedulers by @WenjieDu in #195
• Add learning-rate schedulers for optimizers, and by @WenjieDu in #196
• Releasing v0.1.3; by @WenjieDu in #197

New Contributors

• @vemuribv made their first contribution in #179
• @AugustJW made their first contribution in #176

Full Changelog: v0.1.2...v0.1.3

v0.1.2

In this new version, we mainly fix some bugs, update the docs, and update TSDB changed APIs to keep compatible with it.

What's Changed

• Update logging, and fix the error in the usage example by @WenjieDu in #131
• Fix single-item device list throwing an error by @MaciejSkrabski in #133
• Fix the single-item device list error, update docs by @WenjieDu in #134
• Specifying the version of protobuf<=4.21.12 to fix issue #135 by @WenjieDu in #136
• Update docs and citation info by @WenjieDu in #140
• Fix daily testing by @WenjieDu in #143
• Fixing failed daily testing with MacOS python 3.7, and update docs by @WenjieDu in #144
• Update docs by @WenjieDu in #148
• Fixing the bug in raindrop when running on multiple devices by @WenjieDu in #149
• Updating docs, and fixing the bug in raindrop when running on multiple devices by @WenjieDu in #150
• Fixing the invalid badge of PyPI downloads by @WenjieDu in #152
• Updating stale workflow, and updating docs by @WenjieDu in #154
• Update actions/stale@v8, and fix docs typo by @WenjieDu in #156
• fix: the type hint for function mcar() in comments by @yhzhu99 in #157
• Update the greeting workflow, and update the docs by @WenjieDu in #159
• Updating workflows and docs by @WenjieDu in #162
• Fixing the bug caused by changed tsdb api, updating the docs by @WenjieDu in #166
• Updating docs and releasing v0.1.2 by @WenjieDu in #167

New Contributors

• @yhzhu99 made their first contribution in #157

Full Changelog: v0.1.1...v0.1.2

v0.1.1 is here! And we bring you parallel training on multiple GPUs 🚀

In this new version, we

1. enable PyPOTS to run on multiple CUDA devices to train models in parallel. NOW you can further speed up your model training in PyPOTS 🥳;
2. add the imputation model M-RNN;
3. add the optimizer Adadelta;

What's Changed

• Use relative imports replace absolute ones by @WenjieDu in #117
• Add optimizer Adadelta, and fix issue #116 by @WenjieDu in #118
• Docs auto param type hints by @WenjieDu in #119
• Use sphinx_autodoc_typehints to auto gene type hints for params, and improve docs by @WenjieDu in #120
• Enable parallel training on multi GPUs by @WenjieDu in #122
• skip multi-gpu test if not multi-gpu host by @MaciejSkrabski in #124
• Enable models to run on multiple CUDA devices by @WenjieDu in #125
• Add model MRNN by @WenjieDu in #126
• Add model M-RNN, and update the docs by @WenjieDu in #127
• Fix VaDER failed testing and MRNN failed training on multiple devices by @WenjieDu in #128

Full Changelog: v0.1.0...v0.1.1

v0.1.0

In this new version, we

1. separated each model into a single package, for better standardization and easier management in the library;
2. added model templates, for each task to assist PyPOTS contributors to quickly add new models;
3. isolated optimizers from models,
   1). to separate hyperparameters and avoid putting all of them in one place, which could result in a mess and be not readable;
   2). to provide additional functionalities in the near future, such as learning rate scheduling, etc.
4. Documentations are formally released! 🚀
   Visit PyPOTS website to see them;

What's Changed

• Try to automatically fix the exception ValueError raised by gmm.fit() during VaDER training by @WenjieDu in #62
• Fix the exceptionValueError raised by gmm.fit() during VaDER training, update README, move logo figs, and remove version.py by @WenjieDu in #63
• Add pypots-cli dev by @WenjieDu in #66
• Add pre-commit and enable pre-commit hooks to check code before commtting by @WenjieDu in #68
• Add pre-commit config file and update issue templates by @WenjieDu in #69
• Add unit-test cases for pypots-cli by @WenjieDu in #72
• Fix bugs in the code-coverage report by @WenjieDu in #73
• Add devcontainer config by @WenjieDu in #76
• Tutorial release, new features, and bug fix by @WenjieDu in #75
• Fix dependency error in daily testing by @WenjieDu in #77
• Add PR template, and make the dependencies consistent in all config files by @WenjieDu in #78
• Update READE, and separate the number of trainable parameters with commas by @WenjieDu in #79
• Fix the footer issue on the docs home page, and update the docs by @WenjieDu in #81
• Update the docs with the pages installation and about_us added by @WenjieDu in #82
• Add the documentation pages of FAQ and examples by @WenjieDu in #83
• Enable different model saving strategies with argument model_saving_strategy by @WenjieDu in #84
• Make each model a single package by @WenjieDu in #86
• Fix failed docs building, update the docs, and allow tests for CLI to fail by @WenjieDu in #89
• Fix failed daily-testing workflow, rename filesdataset.py into data.py, and update the docs by @WenjieDu in #91
• Add gtag to the docs site by @WenjieDu in #92
• Update docs by @WenjieDu in #93
• Update the documentation by @WenjieDu in #94
• Fix the corrupted testing badge caused by renaming the workflow by @WenjieDu in #95
• Move tests from dir pypots to the root dir by @WenjieDu in #96
• Fix not shown heading Attributes in docs by @WenjieDu in #98
• Fix not shown heading Attributes in docs, and update the docs by @WenjieDu in #99
• Should use package pypots as the source for code coverage calculation, rather than tests by @WenjieDu in #101
• Update docs, and fix the misdisplayed footer on the page References by @WenjieDu in #103
• Create package pypots.optim to separate optimizers from models by @WenjieDu in #106
• Add package pypots.optim to separate optimizers from models, and update docs by @WenjieDu in #107
• Update the docs for version 0.1 by @WenjieDu in #104
• Update the docs by @WenjieDu in #108
• Update the docs, and refactor the code by @WenjieDu in #109
• Add .coveragerc to omit files in templates by @WenjieDu in #111
• Add model templates, and release v0.1 by @WenjieDu in #110
• Fix the failed auto release workflow with missing dependencies by @WenjieDu in #113

Full Changelog: v0.0.11...v0.1

v0.0.11

PyPOTS v0.0.11 is released now! 🤩

In this new version, we

1️⃣. added issue templates to help our community open issues properly in different categories;
2️⃣. enabled a code-linting workflow applying Flake8;
3️⃣. implemented the training log saving feature, which can help PyPOTS users better know models' performance during the training stage so that now tuning hyper-parameters is easier;
4️⃣. eased dependency management with file setup.cfg;
5️⃣. fixed an emergency bug🐞 "tensors not on the same device" when running on CUDA devices❗️

Also, together with this release, we initialize the development of pypots-cli, a command line interface tool currently to help PyPOTS developers run tests, code linting, and doc generating, and will enable users to run PyPOTS with shell commands in the future.

PyPOTS is growing 🚀. Stay tuned please!

What's Changed

• Merge dev into main to update the documentation and add doc-generating shell scripts by @WenjieDu in #40
• Add issue templates by @WenjieDu in #41
• Decrease testing python version 3.11 to 3.10, and remove fixed dependency versions by @WenjieDu in #43
• Save training logs into tensorboard files, enable setting num_workers of DataLoader, and add typing annotations by @WenjieDu in #42
• Make imputation models val_X_intact and val_indicating_mask be included in input val_set originally by @WenjieDu in #44
• Refactor the code and correct the error path of the pip-requirement file for Daily-testing workflow by @WenjieDu in #45
• Save training logs into tensorboard files by @WenjieDu in #46
• Implemented saving logs into tb file for models, and added test cases for log and model saving by @WenjieDu in #47
• Add the lint-code workflow and pypots-cli to help development by @WenjieDu in #48
• Merge 'dev' into 'main': add the lint-code workflow and pypots-cli by @WenjieDu in #49
• Update limitations on dependency versions by @WenjieDu in #50
• Add version limitations on dependency 'torch' and 'pandas' by @WenjieDu in #51
• Update issue templates and fix the python version error in daily testing workflow by @WenjieDu in #53
• Add issue template config file and fix errors in testing_daily workflow by @WenjieDu in #54
• Use setup.cfg to manage dependencies for different usages by @WenjieDu in #55
• Add CLI tool pypots-cli doc and update the dependencies of the development env by @WenjieDu in #56
• Update documentation configurations and add readthedocs.yml for processing control by @WenjieDu in #60
• Fix the bug of tensors not on the same device when running on CUDA device by @WenjieDu in #59
• Fix the error of non-valid license classifier in setup.py by @WenjieDu in #61

Full Changelog: v0.0.10...v0.0.11