regression.py

from pygrn import grns, problems, evolution, config
import argparse
import os

parser = argparse.ArgumentParser(description='Evolve a GRN for regression')
parser.add_argument('--no-learn', dest='learn', action='store_const',
                    const=False, default=True,
                    help='Turn off learning')
parser.add_argument('--no-evo', dest='evo', action='store_const',
                    const=False, default=True,
                    help='Turn off evolution')
parser.add_argument('--lamarckian', dest='lamarckian', action='store_const',
                    const=True, default=False, help='Lamarckian evolution')
parser.add_argument('--unsupervised', dest='unsupervised', action='store_const',
                    const=True, default=False, help='Unsupervised evolution')
parser.add_argument('--stateful', dest='stateful', action='store_const',
                    const=True, default=False, help='Stateful model')
parser.add_argument('--id', type=str, help='Run id for logging')
parser.add_argument('--model', type=str, help='Model')
parser.add_argument('--ntrain', type=int, default=6*24*60, help='Number of training samples')
parser.add_argument('--ntest', type=int, default=24*60, help='Number of testing samples')
parser.add_argument('--shift', type=int, default=1, help='Data shift')
parser.add_argument('--lag', type=int, default=60, help='Time step for prediction')
parser.add_argument('--nreg', type=int, default=10,
                    help='Number of starting regulatory proteins')
parser.add_argument('--nout', type=int, default=1,
                    help='Number of outputs proteins, unit size for LSTM or SimpleRNN')
parser.add_argument('--batch_size', type=int, default=1,
                    help='Batch size')
parser.add_argument('--seed', type=int, help='Random seed',
                    default=0)
parser.add_argument('--problem', type=str, help='Problem',
                    default='Prediction')
parser.add_argument('--epochs', type=int, help='Number of epochs', default=1)
parser.add_argument('--gens', type=int, help='Number of generations',
                    default=50)
parser.add_argument('--grn_file', type=str, help='Experts from GRN file',
                    default='')
parser.add_argument('--root_dir', type=str, help='Root directory',
                    default='./')

args = parser.parse_args()

log_dir = os.path.join(args.root_dir, 'logs')
grn_dir = os.path.join(args.root_dir, 'grns')
data_dir = os.path.join(args.root_dir, 'data')
log_file = os.path.join(log_dir, 'fits_' + args.id + '.log')

config.START_REGULATORY_SIZE = args.nreg

p = eval('problems.' + args.problem)
p = p(log_file, seed=args.seed, learn=args.learn, epochs=args.epochs,
      data_dir=data_dir, lamarckian=args.lamarckian,
      stateful=args.stateful, model=args.model,
      nout=args.nout, batch_size=args.batch_size)
# p = p(log_file, seed=args.seed, learn=args.learn, epochs=args.epochs,
#       data_dir=data_dir, lamarckian=args.lamarckian,
#       unsupervised=args.unsupervised, stateful=args.stateful,
#       model=args.model, ntrain=args.ntrain, ntest=args.ntest,
#       shift=args.shift, lag=args.lag)

newgrn = lambda: grns.DiffGRN()
if args.evo:
    grneat = evolution.Evolution(p, newgrn, run_id=args.id,
                                 grn_dir=grn_dir, log_dir=log_dir)
    grneat.run(args.gens)
else:
    for i in range(20):
        grn = newgrn()
        if args.grn_file:
            with open(args.grns, 'r') as f:
                grns = f.readlines()
                grn.from_str(grns[-1])
        else:
            grn.random(p.nin, p.nout, args.nreg)
        p.generation_function(None, i)
        p.eval(grn)