What's Changed

GenNet interpret: adds the interpretation module with six interpretation methods
- get_weight_scores: uses the weights to calculate the importance of each feature and node
- DeepExplain: uses the gradient (see DeepExplain) to calculate the importance
- RLIPP: uses logistic regression with signals to and from the node to calculate a measure of non-linearity for all nodes
- NID: Finds interacting features based on the features with the strongest weights
- DFIM: perturbs each input (or N inputs in the order of importance), and tracks which other features change importance to find interacting features
- PathExplain: Uses the Expected Hessian to find interacting features
Bugfixes for converting plink2 files
Script for converting toplogy to npz matrices
Add one-hot encoding support
Adds multiple filters. LocallyDirected1D now supports multiple filters akin to channels or feature maps in CNNs

Assets 2

03 Aug 09:47

Release GenNet 1.5

A to Z tutorial
Pytest and automatic integration
Topology allows you now to create a pathway network with create_pathway_network
Continue and resume training
Specify your epoch length for large datasets
There are now options for multiprocessing, mixed_precision
Add suffix for better naming
Genotype_path option so that it does not have to be in the same folder as the other files
Revamped training for regression problems

Bug fixes

Assets 2

01 Aug 14:56

Release GenNet 1.4

Minor improvements.

Added submodule toplogy.
Improved plotting with and without chr information
Can now use .npz files instead of topology.csv files. (make sure that all the maks are in the same folder and that the output size corresponds to the input size of the next mask.

Assets 2