A Leap Motion sign language classifier

slang.py

import sys
import getopt
import csv
import random

def parse_csv(file):
    with open(file) as csvfile:
        csvreader = csv.reader(csvfile)
        data = []
        target = []
        for row in csvreader:
            target.append(row.pop())
            data.append([float(coord) for coord in row])
    return data, target

def train_knn(training):
    d,t = parse_csv(training)

    cut = int(len(d) * 0.7)
    training_set    = d[cut:]
    labeling_set    = t[cut:]
    validation_set  = d[:cut]

    from sklearn import neighbors
    knn = neighbors.KNeighborsClassifier()
    knn.fit(training_set, labeling_set)
    return knn


def predict(data, training):
    classifier = train_knn(training)
    klass = classifier.predict(data)
    return klass


def usage():
    print 'Usage:'
    print sys.argv[0], '--predict <DATA> --from <TRAINING_SET>'


def main(argv):
    training, predictable = '', ''

    try:
        opts, args = getopt.getopt(argv, "htf:p:", ["help", "test", "from=", "predict="])
    except getopt.GetoptError, e:
        print str(e)
        usage()
        sys.exit(2)
    for opt, arg in opts:
        if opt in ("-h", "--help"):
            usage()
            sys.exit()
        elif opt in ("-t", "--test"):
            data = '0.516162, 0.856276, -0.0192088, -0.659962, 0.411919, 0.628309, 0.545918, -0.311632, 0.777727, 0.402093, 0.9001, -0.167755, -0.487101, 0.365433, 0.793216, 0.775276, -0.237233, 0.585378, 0.402093, 0.9001, -0.167755, -0.439081, 0.350339, 0.827328, 0.803448, -0.259005, 0.536085, 0.402093, 0.9001, -0.167755, -0.330934, 0.313705, 0.889984, 0.8537, -0.302341, 0.424012, 0.940846, -0.0730333, -0.330869, 0.120909, 0.984572, 0.126485, 0.316527, -0.159008, 0.935162, 0.944734, -0.0750178, -0.319138, 0.104466, 0.991608, 0.0761574, 0.310747, -0.105288, 0.944643, 0.944734, -0.0750178, -0.319138, 0.0899903, 0.995415, 0.0324095, 0.315244, -0.0593377, 0.947154, 0.944734, -0.0750178, -0.319138, 0.0784608, 0.996915, -0.00207365, 0.318309, -0.0230808, 0.947706, 0.94047, -0.260613, -0.218166, 0.285149, 0.954321, 0.0892271, 0.184946, -0.146125, 0.971824, 0.953749, -0.273267, -0.125253, 0.276753, 0.960878, 0.0109837, 0.117351, -0.0451397, 0.992064, 0.953749, -0.273267, -0.125253, 0.270764, 0.961937, -0.0369247, 0.130575, 0.00130301, 0.991438, 0.953749, -0.273267, -0.125253, 0.265647, 0.961204, -0.0742896, 0.140694, 0.0375806, 0.98934, 0.933247, -0.35047, -0.0788773, 0.357363, 0.928118, 0.104346, 0.0366372, -0.125569, 0.991408, 0.933495, -0.351402, -0.0714364, 0.356172, 0.931711, 0.0711032, 0.0415723, -0.0918181, 0.994908, 0.933495, -0.351402, -0.0714364, 0.353771, 0.935041, 0.0233504, 0.0585906, -0.0470696, 0.997172, 0.933495, -0.351402, -0.0714364, 0.35129, 0.936154, -0.0145415, 0.0719854, -0.0115205, 0.997339, 0.859253, -0.51115, 0.0202481, 0.49999, 0.847544, 0.177984, -0.108137, -0.142809, 0.983825, 0.847765, -0.520984, 0.0993472, 0.401102, 0.752342, 0.522589, -0.347003, -0.403184, 0.846777, 0.847765, -0.520984, 0.0993472, 0.521307, 0.85301, 0.0247423, -0.0976346, 0.0308148, 0.994745, 0.847765, -0.520984, 0.0993472, 0.508154, 0.744228, -0.433478, 0.151898, 0.417972, 0.895671'
            from subprocess import call
            call(["python", sys.argv[0], "-f", "training.csv", "-p", data])
        elif opt in ("-f", "--from"):
            training = arg
        elif opt in ("-p", "--predict"):
            predictable = arg.split(',')

    if(len(predictable) > 0):
        prediction = predict(predictable, training)
        print "Class: ", prediction
    else:
        return -1

if __name__ == "__main__":
    main(sys.argv[1:])

In action:

➜ slang python slang.py --from training.csv --predict '0.656128, 0.536473, 0.530747, -0.210324, -0.545438, 0.811333, 0.724748, -0.643967, -0.245044, 0.745088, 0.447858, 0.494234, -0.0857279, -0.670564, 0.736881, 0.661434, -0.591411, -0.461235, 0.745088, 0.447858, 0.494234, -0.481741, -0.151113, 0.863186, 0.46127, -0.881243, 0.103159, 0.745088, 0.447858, 0.494234, -0.58217, 0.0751144, 0.80959, 0.325457, -0.890944, 0.316696, 0.664667, 0.519156, -0.537303, 0.404042, 0.355157, 0.842979, 0.628465, -0.777393, 0.0262997, 0.671776, 0.510275, -0.536969, -0.575021, 0.816203, 0.0562469, 0.466977, 0.270983, 0.841725, 0.671776, 0.510275, -0.536969, -0.662847, 0.0904882, -0.743267, -0.330681, 0.855238, 0.399022, 0.671776, 0.510275, -0.536969, -0.435463, -0.314371, -0.84353, -0.59924, 0.800494, 0.0110191, 0.66478, 0.341151, -0.664592, 0.524295, 0.42066, 0.740378, 0.532148, -0.840631, 0.100782, 0.670273, 0.3324, -0.663509, -0.467466, 0.883515, -0.0296136, 0.576377, 0.330017, 0.747582, 0.670273, 0.3324, -0.663509, -0.718938, 0.069146, -0.691626, -0.184018, 0.9406, 0.285322, 0.670273, 0.3324, -0.663509, -0.55395, -0.370861, -0.745387, -0.493837, 0.867163, -0.0644449, 0.694417, 0.181039, -0.696427, 0.587147, 0.416946, 0.69384, 0.415985, -0.890719, 0.183238, 0.687182, 0.196285, -0.699466, -0.3608, 0.927886, -0.09408, 0.630558, 0.317018, 0.708446, 0.687182, 0.196285, -0.699466, -0.723782, 0.101997, -0.682448, -0.0626105, 0.975228, 0.212158, 0.687182, 0.196285, -0.699466, -0.613871, -0.358011, -0.703556, -0.388514, 0.912853, -0.125525, 0.631272, 0.00533413, -0.775543, 0.722436, 0.359687, 0.590518, 0.282103, -0.933058, 0.223207, 0.639684, -0.0234884, -0.768279, 0.00313799, 0.999604, -0.0279479, 0.768631, 0.015467, 0.639505, 0.639684, -0.0234884, -0.768279, -0.716368, 0.34408, -0.606981, 0.278607, 0.938646, 0.203276, 0.639684, -0.0234884, -0.768279, -0.755755, -0.201453, -0.623098, -0.140136, 0.979217, -0.146617'
Class: [" 'y'"]