Skip to content

jondeaton/HumanProteinAtlas

Repository files navigation

HumanProteinAtlas

CS229 Final Project

Team members
  • Jon Deaton
  • Robert Neff
  • Luke Sturm

Usage

Extract features and cluster features

python -m cell_clustering.cluster \
    --features-dir /path/to/features/directory \
    --model-file  /path/to/model/file \
    --assignments-file /path/to/assignments/file

To train deep model

python -m deep_model.train

evaluate the deep model

python -m deep_model.evaluate \
    --save-path /path/to/model_outputs \
    --model model_checkpoint-0.meta \
    --output outputs

To create training, test, and validation sets (this has already been done and should not need to be done again)

python -m partitions.create --input ~/Datasets/HumanProteinAtlas --num-test 1500 --num-validation 1500