IterPert
Project description
Iterative Perturb-seq
Install
pip install iterpert
API interface
from iterpert.iterpert import IterPert
strategy = 'IterPert' # choose from 'Random', 'BALD', 'BatchBALD', 'BAIT', 'ACS-FW', 'Core-Set', 'BADGE', 'LCMD', 'IterPert'
interface = IterPert(weight_bias_track = True,
exp_name = strategy,
device = 'cuda:0',
seed = 1)
path = '/home/huangk28/scratch/perturb_seq_data/gears_data/'
interface.initialize_data(path = path,
dataset_name='replogle_k562_essential_1000hvg',
batch_size = 256)
interface.initialize_model(epochs = 20, hidden_size = 64)
interface.initialize_active_learning_strategy(strategy = strategy)
interface.start(n_init_labeled = 100, n_round = 5, n_query = 100)
Reproduce experiments
Please refer to reproduce_repo
directory to reproduce each experiment. Notably, reproduce_script.sh
contains sh files to generate all experiments. figX.ipynb
is the notebook that produces the figures.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
iterpert-0.0.1.tar.gz
(80.7 kB
view hashes)