Version 1.0.6 minor release

Qiime2 diagnostic summaries are enabled, allowing for loss curves to be visualized in addition to comparisons to null models.

Minor release

Fixed issue with batch size not being exposed. Now with the batch-size parameter available, it is easier to tune to speed up calculations, which is important for GPU compute.

Minor release

This release includes an option to equalize biplot vectors (thanks to @mstrazar ).

Dependency issues due to pandas and tensorflow have been temporarily fixed.

Version 1.0.3 (Holiday edition)

In this release, we have two minor updates, namely

• Setting the default learning rate to 1e-5
• Pinning tensorflow to any version below 2.0 in both pypi and conda

Version 1.0.2

Heatmaps have a row center option, allow for users to specify if they want to center the conditional probabilities around the average row.

Version 1.0.1

In this release, the ranks have been transposed, enabling easier viewing and sorting of metabolites in downstream spreadsheet programs and qiime metadata tabulate

Version 1 release

This marks the version 1 release of mmvec.

Major contributions to this release include paired heatmap visualizations that can be accessed through qiime mmvec paired-heatmaps.

Furthermore, the tutorials have been updated to reflect the changes made to the heatmaps.
See changelog for other notes on bug fixes.

Version 0.6.0

This is an intermediate release, enabling the production of heatmaps to visualize ranks and tighter compatibility between the standalone CLI and the qiime2 CLI. Now ranks can be directly imported as qiime2 objects.

In addition, the package has been renamed from rhapsody to mmvec. Furthermore the command mmvec has been changed to paired-omics.

This will be the last intermediate release before version 1.

Improvements to standalone interface

The standalone command line interface has been fixed and simplified. Now, all of the output files have readable labels.

Beta release

This marks the beta release of rhapsody.

Bugs have been squashed, tests have been added, and documentation has been added.
This includes tutorials using the cystic fibrosis dataset.