Federated version of PLIER.
For more information on Flower, we refer the reader to Flower's webpage.
Install the python environment by running conda env create -f flower.yml. This will ensure the packages versions work correctly.
In the . directory, use flwr run to run a local simulation of the federated process:
flwr run . --streamThis will load the data inside data/client_* and start the simulated federation. This version uses the complete gene data in data/server/ and splits it in options.num-supernodes parts (the number of clients). At the end of the simulation, 2 images will be created:
figures/errorY_curve.png: the curve of the difference between original Y matrix and reconstructed one.
figures/reconstruction_plot.png: A spearman correlation curve comparing the federated version to the centralized correlation.

