• Adding support for multiple targets in the TimeSeriesDataSet (#199) and amended tutorials.
• Temporal fusion transformer and DeepAR with support for multiple tagets (#199)
• Check for non-finite values in TimeSeriesDataSet and better validate scaler argument (#220)

• TimeSeriesDataSet's y of the dataloader is a tuple of (target(s), weight) - potentially breaking for model or metrics implementation Most implementations will not be affected as hooks in BaseModel and MultiHorizonMetric were modified.

• Fixed autocorrelation for pytorch 1.7 (#220)
• Ensure reproducibility by replacing python set() with dict.fromkeys() (#221)

• jdb78
• JustinNeumann

• Tutorial on how to implement a new architecture covering basic and advanced use cases (#188)
• Additional and improved documentation - particularly of implementation details (#188)

• Moved multiple private methods to public methods (particularly logging) (#188)
• Moved get_mask method from BaseModel into utils module (#188)
• Instead of using label to communicate if model is training or validating, using self.training attribute (#188)
• Using sample((n,)) of pytorch distributions instead of deprecated sample_n(n) method (#188)

• Beta distribution loss for probabilistic models such as DeepAR (#160)

• BREAKING: Simplifying how to apply transforms (such as logit or log) before and after applying encoder. Some transformations are included by default but a tuple of a forward and reverse transform function can be passed for arbitrary transformations. This requires to use a transformation keyword in target normalizers instead of, e.g. log_scale (#185)

• Incorrect target position if len(static_reals) > 0 leading to leakage (#184)
• Fixing predicting completely unseen series (#172)

• jdb78
• JakeForsey

• Using GRU cells with DeepAR (#153)

• GPU fix for variable sequence length (#169)
• Fix incorrect syntax for warning when removing series (#167)
• Fix issue when using unknown group ids in validation or test dataset (#172)
• Run non-failing CI on PRs from forks (#166, #156)

• Improved model selection guidance and explanations on how TimeSeriesDataSet works (#148)
• Clarify how to use with conda (#168)

• jdb78
• JakeForsey

• DeepAR by Amazon (#115)
  • First autoregressive model in PyTorch Forecasting
  • Distribution loss: normal, negative binomial and log-normal distributions
  • Currently missing: handling lag variables and tutorial (planned for 0.6.1)
• Improved documentation on TimeSeriesDataSet and how to implement a new network (#145)

• Internals of encoders and how they store center and scale (#115)

• Update to PyTorch 1.7 and PyTorch Lightning 1.0.5 which came with breaking changes for CUDA handling and with optimizers (PyTorch Forecasting Ranger version) (#143, #137, #115)

• jdb78
• JakeForesey

• Fix issue where hyperparameter verbosity controlled only part of output (#118)
• Fix occasional error when .get_parameters() from TimeSeriesDataSet failed (#117)
• Remove redundant double pass through LSTM for temporal fusion transformer (#125)
• Prevent installation of pytorch-lightning 1.0.4 as it breaks the code (#127)
• Prevent modification of model defaults in-place (#112)

• Hyperparameter tuning with optuna to tutorial
• Control over verbosity of hyper parameter tuning

• Interpretation error when different batches had different maximum decoder lengths
• Fix some typos (no changes to user API)

This release has only one purpose: Allow usage of PyTorch Lightning 1.0 - all tests have passed.

• Additional checks for TimeSeriesDataSet inputs - now flagging if series are lost due to high min_encoder_length and ensure parameters are integers
• Enable classification - simply change the target in the TimeSeriesDataSet to a non-float variable, use the CrossEntropy metric to optimize and output as many classes as you want to predict

• Ensured PyTorch Lightning 0.10 compatibility
  • Using LearningRateMonitor instead of LearningRateLogger
  • Use EarlyStopping callback in trainer callbacks instead of early_stopping argument
  • Update metric system update() and compute() methods
  • Use trainer.tuner.lr_find() instead of trainer.lr_find() in tutorials and examples
• Update poetry to 1.1.0

• Removed attention to current datapoint in TFT decoder to generalise better over various sequence lengths
• Allow resuming optuna hyperparamter tuning study

• Fixed inconsistent naming and calculation of encoder_lengthin TimeSeriesDataSet when added as feature

• jdb78

• Backcast loss for N-BEATS network for better regularisation
• logging_metrics as explicit arguments to models

• MASE (Mean absolute scaled error) metric for training and reporting
• Metrics can be composed, e.g. 0.3* metric1 + 0.7 * metric2
• Aggregation metric that is computed on mean prediction over all samples to reduce mean-bias

• Increased speed of parsing data with missing datapoints. About 2s for 1M data points. If numba is installed, 0.2s for 1M data points
• Time-synchronize samples in batches: ensure that all samples in each batch have with same time index in decoder

• Improved subsequence detection in TimeSeriesDataSet ensures that there exists a subsequence starting and ending on each point in time.
• Fix min_encoder_length = 0 being ignored and processed as min_encoder_length = max_encoder_length

• jdb78
• dehoyosb

• More tests driving coverage to ~90%
• Performance tweaks for temporal fusion transformer
• Reformatting with sort
• Improve documentation - particularly expand on hyper parameter tuning

• Fix PoissonLoss quantiles calculation
• Fix N-Beats visualisations

• Calculating partial dependency for a variable
• Improved documentation - in particular added FAQ section and improved tutorial
• Data for examples and tutorials can now be downloaded. Cloning the repo is not a requirement anymore
• Added Ranger Optimizer from pytorch_ranger package and fixed its warnings (part of preparations for conda package release)
• Use GPU for tests if available as part of preparation for GPU tests in CI

• BREAKING: Fix typo “add_decoder_length” to “add_encoder_length” in TimeSeriesDataSet

• Fixing plotting predictions vs actuals by slicing variables

Fix bug where predictions were not correctly logged in case of decoder_length == 1.

• Add favicon to docs page

Update build system requirements to be parsed correctly when installing with pip install https://github.com/jdb78/pytorch-forecasting/

• Add tests for MacOS
• Automatic releases
• Coverage reporting

This release improves robustness of the code.

Fixing bug across code, in particularly

• Ensuring that code works on GPUs
• Adding tests for models, dataset and normalisers
• Test using GitHub Actions (tests on GPU are still missing)

Extend documentation by improving docstrings and adding two tutorials.

• Basic tests for data and model (mostly integration tests)
• Automatic target normalization
• Improved visualization and logging of temporal fusion transformer
• Model bugfixes and performance improvements for temporal fusion transformer

• Metrics are reduced to calculating loss. Target transformations are done by new target transformer