API reference#
bulkdgd consists of several packages:
analysis, containing utilities to analyze the data produced by the bulkdgd model.
core, containing the core components of the bulkdgd model and the model itself.
genes, containing utilities to create customized lists of genes to use with the bulkdgd model.
ioutil, containing utilities for loading and saving files and pre-process data.
plotting, containing plotting utilities.
recount3, containing utilities to interact with the Recount3 platform.
reproducibility, containing
bulkdgd.reproducibility.set_seeds(), which seeds the generators a run draws from. It has to be called before the model is built, since building it already draws the decoder’s weights.
Some functions and methods use dictionary-based configurations for several tasks, such as setting up the bulkdgd model, finding the best representations for a new set of samples and training the bulkdgd model. More detailed descriptions of such configurations are available here:
configuration used to set the optimization scheme when finding representations for a new set of samples.