def runActualTest(points): actual = True print 'RUNNING PREDICTION: ', ['Actual' if actual else 'Test'] points = '33.46503917 26.76995952 85.97458381 82.61217489 59.33400217 49.29506539'.split( ) #points = '5 51 4 73 55 36'.split() points = map(float, points) #points = np.array(zip(points[::2],points[1::2])) print points testTimes = pd.read_csv('test_times.csv') tags = generatetest.listTags() if actual: trainTags = tags[0:100] else: tags = tags[0:100] trainTags = tags[0:10] print 'Training Tags: ', trainTags print 'Test Tags: ', tags optFunc = makeOptFunc(testTimes, trainTags, tags, visualise=True) result = optFunc(np.array([points])) print result quit()
def runActualTest(points): actual = True print 'RUNNING PREDICTION: ', ['Actual' if actual else 'Test'] points = '33.46503917 26.76995952 85.97458381 82.61217489 59.33400217 49.29506539'.split() #points = '5 51 4 73 55 36'.split() points = map(float, points) #points = np.array(zip(points[::2],points[1::2])) print points testTimes = pd.read_csv('test_times.csv') tags = generatetest.listTags() if actual: trainTags = tags[0:100] else: tags = tags[0:100] trainTags = tags[0:10] print 'Training Tags: ', trainTags print 'Test Tags: ', tags optFunc = makeOptFunc(testTimes, trainTags, tags, visualise=True) result = optFunc(np.array([points])) print result quit()
print 'Training Tags: ', trainTags print 'Test Tags: ', tags optFunc = makeOptFunc(testTimes, trainTags, tags, visualise=True) result = optFunc(np.array([points])) print result quit() if __name__ == '__main__': if len(sys.argv) > 1: runActualTest(sys.argv[1:]) testTimes = pd.read_csv('test_times.csv') tags = generatetest.listTags()[0:100] trainTags, testTags = generatetest.splitTags(tags, proportion=0.1) print 'Training Tags: ', trainTags print 'Test Tags: ', tags acquisition_par = 0.01 max_iter = 10 bounds = [(0, 100)] * 6 optFunc = makeOptFunc(testTimes, trainTags, tags, visualise=True) bOpt = GPyOpt.methods.BayesianOptimization(optFunc, bounds=bounds, acquisition='LCB', acquisition_par=acquisition_par) bOpt.run_optimization(max_iter,
print 'Training Tags: ', trainTags print 'Test Tags: ', tags optFunc = makeOptFunc(testTimes, trainTags, tags, visualise=True) result = optFunc(np.array([points])) print result quit() if __name__ == '__main__': if len(sys.argv) > 1: runActualTest(sys.argv[1:]) testTimes = pd.read_csv('test_times.csv') tags = generatetest.listTags()[0:100] trainTags, testTags = generatetest.splitTags(tags, proportion=0.1) print 'Training Tags: ', trainTags print 'Test Tags: ', tags acquisition_par = 0.01 max_iter = 10 bounds = [(0, 100)] * 6 optFunc = makeOptFunc(testTimes, trainTags, tags, visualise=True) bOpt = GPyOpt.methods.BayesianOptimization(optFunc, bounds=bounds, acquisition='LCB', acquisition_par=acquisition_par)