predict.py

__author__ = 'dudevil'

import sys
import os.path
import argparse
import numpy as np
import pandas as pd
import theano
import theano.tensor as T
from lasagne.utils import floatX
from utils import load_network, get_predictions
from load_dataset import DataLoader


BATCH_SIZE = 64
IMAGE_SIZE = 128

def save_submission(predictions, filenames, n=1):
    #assert(len(predictions) == len(filenames))
    names = [os.path.splitext(os.path.basename(image))[0] for image in filenames]
    dfr = pd.DataFrame(predictions[:len(filenames)], index=names)
    dfr.to_csv(os.path.join("data", "submissions", "submission_%s.csv" % n))

if __name__ == "__main__":
    parser = argparse.ArgumentParser()
    parser.add_argument("-n",
                        "--network",
                        type=str,
                        default=os.path.join("data", "tidy", "net.pickle"),
                        help="Path to the pickled network file")
    parser.add_argument('--proba', dest='proba', action='store_const',
                        const=1, default=0,
                        help='flag to predict probabilities rather than hard targets')

    args = parser.parse_args()
    netfile = args.network

    print("Loading saved network...")
    if not os.path.isfile(netfile):
        print("No such file: %s" % netfile)
        exit()
    try:
        network, output = load_network(netfile)
    except Exception, e:
        print("Could not load network: %s" % e)

    print("Loading test dataset...")
    # load test data chunk
    dl = DataLoader(image_size=IMAGE_SIZE,
                    normalize=True,
                    batch_size=64,
                    parallel=False,
                    train_path="train/trimmed",
                    test_path=os.path.join("test", "trimmed"))
    test_filenames = dl.test_images
    n_predictions = len(test_filenames)
    print("Compiling theano functions...")
    # set up symbolic variables
    X = T.tensor4('X')
    X_batch = T.tensor4('X_batch')
    batch_index = T.iscalar('batch_index')

    #pred = T.iround(output.get_output(X_batch, deterministic=True))
    if args.proba:
        predict = theano.function(
            [theano.Param(X_batch)],
            output.get_output(X_batch, deterministic=True),
            givens={
                X: X_batch
                },
            )
    else:
        predict = theano.function(
            [theano.Param(X_batch)],
            T.gt(output.get_output(X_batch, deterministic=True), 0.5),
            givens={
                X: X_batch
                },
            )
    print("Predicting...")
    predictions = []

    for test_chunk in dl.test_gen():
        preds = predict(test_chunk)
        predictions.append(preds)
        sys.stdout.write("progress: %d %%\r" % (len(predictions) * BATCH_SIZE * 100. / n_predictions))
        sys.stdout.flush()
    print("Saving predictions")
    predictions = np.vstack(predictions)
    if not args.proba:
        predictions = get_predictions(predictions)
        save_submission(predictions.flatten(), test_filenames, n=23)
    else:
        save_submission(predictions, test_filenames, n='19')