ar.sort() flag = np.concatenate(([True], ar[1:] != ar[:-1])) return ar[flag] ############################################################################### # end sklearn (http://scikit-learn.org/) ############################################################################### if __name__ == '__main__': import UCRDatabase as ucr import pylab as pl # read data set from UCR database (testLabels, testSet), (trainLabels, trainSet) = ucr.read(3) rho = float(len(testSet)) / (len(trainSet) + len(testSet)) # union of labels and items labels, items = ucr.merge(testLabels, testSet, trainLabels, trainSet) # check histograms sss = StratifiedShuffleSplit(labels, 15, test_size=rho, random_state=0) for train_index, test_index in sss: print("TRAIN:", train_index, "TEST:", test_index) pl.hist(labels) pl.hist(labels[train_index]) pl.show()
ar.sort() flag = np.concatenate(([True], ar[1:] != ar[:-1])) return ar[flag] ############################################################################### # end sklearn (http://scikit-learn.org/) ############################################################################### if __name__ == '__main__': import UCRDatabase as ucr import pylab as pl # read data set from UCR database (testLabels, testSet), (trainLabels, trainSet) = ucr.read(3) rho = float(len(testSet))/(len(trainSet)+len(testSet)) # union of labels and items labels, items = ucr.merge(testLabels, testSet, trainLabels, trainSet) # check histograms sss = StratifiedShuffleSplit(labels, 15, test_size=rho, random_state=0) for train_index, test_index in sss: print("TRAIN:", train_index, "TEST:", test_index) pl.hist(labels) pl.hist(labels[train_index])
print errors[-1] errs = list(sorted(errors, key=lambda x: x[0][0])) # pick only the best and take parameter in the "middle" best = filter(lambda (x, y): x[0] == errs[0][0][0], errs) best.sort(key=lambda (x, y): y[2]) best = best[len(best) / 2] return best, errs if __name__ == "__main__": import UCRDatabase as ucr import sys for number in [int(sys.argv[1])]: # read data set from UCR database (testLabels, testSet), (trainLabels, trainSet) = ucr.read(number) # z-normalize testSet = np.array(map(ucr.znormalize, testSet)) trainSet = np.array(map(ucr.znormalize, trainSet)) best_dtw = learn_cdtw(trainLabels, trainSet, False)[0] best_gem = learn_gem(trainLabels, trainSet, False, False)[0] print "dtw params", best_dtw print "gem params", best_gem
print errors[-1] errs = list(sorted(errors, key=lambda x: x[0][0])) # pick only the best and take parameter in the "middle" best = filter(lambda (x, y): x[0]==errs[0][0][0], errs) best.sort(key=lambda (x, y): y[2]) best = best[len(best)/2] return best, errs if __name__ == "__main__": import UCRDatabase as ucr import sys for number in [int(sys.argv[1])]: # read data set from UCR database (testLabels, testSet), (trainLabels, trainSet) = ucr.read(number) # z-normalize testSet = np.array(map(ucr.znormalize, testSet)) trainSet = np.array(map(ucr.znormalize, trainSet)) best_dtw = learn_cdtw(trainLabels, trainSet, False)[0] best_gem = learn_gem(trainLabels, trainSet, False, False)[0] print "dtw params", best_dtw print "gem params", best_gem