Scaden v1.1.2

Bugfix release

This release fixes a very small, but annoying bug that I introduced recently while trying to reduce the memory footprint.

Scaden v1.1.1

Minor bugfix release

• Fixed bugs in scaden model definition [#88]
• removed installation instructions for bioconda as not functional at the moment [#86]
• Fixed bug in scaden example [#85]

Scaden v1.1.0

This release drastically reduces the memory usage of Scaden when simulating data from a large number of datasets. There are also some other smaller, new features that will improve the usability. Have a look at the changelog below for a detailed list.

Improvements

• Reduced memory usage of scaden simulate significantly by performing simulation for one dataset at a time.
• Using .h5ad format to store simulated data
• Allow reading data in .h5ad format for improved performance (courtesy of @eboileau)
• Improved logging and using rich progress bar for training
• Gene subsetting is now done only when merging datasets, which will allow to generate different combinations
  of simulated datasets
• Added scaden merge command which allows merging of previously created datasets

Scaden v1.0.2

Improvements

• improved logging
• warn user about unavailable datasets

This release improves Scaden's logging functionality by introducing colored output with the rich library, making it easier to follow the logging output.
Most importantly, it adds a warning if a user accidentally specifies a dataset with --train_datasets which is not actually part of the training data.

Scaden bugfix release v1.0.1

This release fixes a bug related to parsing input dataset. Previously, Scaden would simply split files by underscore and then take
the first element of this split as the name. Now, the pattern is removed instead, making the use of names with multiple underscores possible.

This release fixes issue #66

Scaden v1.0.0

Release of Scaden v1.0.0

Improvements

• switched to Tensorflow 2 Keras API
• added Scaden example functionality
• more robust data parsing

Bug fixes

• removed bug with input path for scaden simulate
• removed typos

Scaden v0.9.6

This release fixes the following issues:

• Docker image now uses pip to install Scaden and not bioconda
• Better help message for scaden simulate
• removed dependencies for scanpy, scipy and seaborn

Scaden v0.9.5

This release brings some major improvements, most importantly the scaden simulate command. Other improvements are:

• added --seed parameter to allow reproducible Scaden runs
• changed CLI calling
• added more error messages
• code restructuring

Scaden v0.9.2

Release of Scaden version 0.9.2

Includes minor changes in the code, among others improved RAM usage
and some resolved bugs.

Scaden v0.9.0

This marks the initial release of Scaden, as used in Deep-learning-based cell composition analysis from tissue expression profiles.

It contains all functionality for deconvolution, but lacks some additional features that are planned for future releases. This released is published on pip and Bioconda (Scaden v0.9.0)