def test_preprocess():
x = np.array([np.random.rand(28, 28)])
y = np.array([2])
x_train_pp, y_train_pp, x_test_pp, y_test_pp = train.preprocess(x, y, x, y)
assert y_train_pp.shape == (1, 10)
assert y_test_pp.shape == (1, 10)
assert x_train_pp.shape == (1, 28, 28, 1)
assert x_test_pp.shape == (1, 28, 28, 1)
def run_test():
BS = 128
print('Loading and preprocessing test data...')
mean, std = Learner.load_meanstd()
imgs_test = load_test_data()
imgs_test = preprocess(imgs_test)
imgs_test = imgs_test.astype('float32')
imgs_test -= mean
imgs_test /= std
print('Loading saved weights...')
model = get_unet(Adam(0.001))
print ('Loading weights from %s' % Learner.best_weight_path)
model.load_weights(Learner.best_weight_path)
print ('Augment')
alen, dlen = len(transforms), len(imgs_test)
test_x = np.ndarray((alen, dlen, 1, img_rows, img_cols), dtype=np.float32)
for i in range(dlen):
for j, transform in enumerate(transforms):
test_x[j,i] = transform['do'](imgs_test[i].copy())
#
print('Predicting masks on test data...')
outputs = []
asis_res = model.predict(imgs_test, batch_size=BS, verbose=1)
outputs.append(asis_res)
for j, transform in enumerate(transforms):
t_y = model.predict(test_x[j], batch_size=BS, verbose=1)
outputs.append(t_y)
#
print('Analyzing')
test_masks = np.ndarray((dlen, 1, img_rows, img_cols), dtype=np.float32)
test_probs = np.ndarray((dlen, ), dtype=np.float32)
for i in range(dlen):
masks = np.ndarray((alen+1, 1, img_rows, img_cols), dtype=np.float32)
probs = np.ndarray((alen+1, ), dtype=np.float32)
for j, t_y in enumerate(outputs):
mask, prob = t_y[0][i], t_y[1][i]
if j:
mask = transforms[j-1]['undo'](mask)
masks[j] = mask
probs[j] = prob
#
test_masks[i] = np.mean(masks, 0)
test_probs[i] = np.mean(probs)
print('Saving ')
np.save(Learner.test_mask_res, test_masks)
np.save(Learner.test_mask_exist_res, test_probs)
def generate_submission():
# Load test images and preprocess for conv net.
print('Loading and processing test images')
imgs_test = DataManager.load_test_data()
total = imgs_test.shape[0]
imgs = np.ndarray(
(total, 1, DataManager.IMG_TARGET_ROWS, DataManager.IMG_TARGET_ROWS),
dtype=np.uint8)
i = 0
for img in imgs_test:
imgs[i] = preprocess(img)
i += 1
print('Loading network')
model = build_model()
model.load_weights('./results/net.hdf5')
print('Generating predictions')
masks, has_masks = model.predict(imgs, verbose=1)
ids = []
rles = []
for i in range(total):
# Zero out masks when there is no-nerve pred.
if has_masks[i, 0] < 0.5:
masks[i, 0] *= 0.
mask = post_process_mask(masks[i, 0])
rle = run_length_enc(mask)
rles.append(rle)
ids.append(i + 1)
if i % 100 == 0:
print('{}/{}'.format(i, total))
first_row = 'img,pixels'
file_name = 'results/submission_{}.csv'.format(str(datetime.now()))
with open(file_name, 'w+') as f:
f.write(first_row + '\n')
for i in range(total):
s = str(ids[i]) + ',' + rles[i]
f.write(s + '\n')
def run_test():
BS = 256
print('Loading and preprocessing test data...')
mean, std = Learner.load_meanstd()
imgs_test = load_test_data()
# imgs_test = imgs_test[:100]
# print ('test')
imgs_test = preprocess(imgs_test)
imgs_test = imgs_test.astype('float32')
imgs_test -= mean
imgs_test /= std
print('Augment')
alen, dlen = len(transforms), len(imgs_test)
test_x = np.ndarray((alen, dlen, 1, img_rows, img_cols), dtype=np.float32)
for i in xrange(dlen):
for j, transform in enumerate(transforms):
test_x[j, i] = transform['do'](imgs_test[i].copy())
#
kfold = 6
kfold_masks, kfold_prob = [], []
for _iter in xrange(kfold):
print('Iter=%d, Loading saved weights...' % _iter)
model = get_unet(Adam(0.001))
filepath = Learner.best_weight_path + '_%d.fold' % _iter
print('Loading weights from %s' % filepath)
model.load_weights(filepath)
#
print('Predicting masks on test data...')
outputs = []
asis_res = model.predict(imgs_test, batch_size=BS, verbose=1)
outputs.append(asis_res)
for j, transform in enumerate(transforms):
t_y = model.predict(test_x[j], batch_size=BS, verbose=1)
outputs.append(t_y)
#
print('Analyzing')
test_masks = np.ndarray((dlen, 1, img_rows, img_cols),
dtype=np.float32)
test_probs = np.ndarray((dlen, ), dtype=np.float32)
for i in xrange(dlen):
masks = np.ndarray((alen + 1, 1, img_rows, img_cols),
dtype=np.float32)
probs = np.ndarray((alen + 1, ), dtype=np.float32)
for j, t_y in enumerate(outputs):
mask, prob = t_y[0][i], t_y[1][i]
if j:
mask = transforms[j - 1]['undo'](mask.copy())
masks[j] = mask
probs[j] = prob
#
test_masks[i] = np.mean(masks, 0)
test_probs[i] = np.mean(probs)
kfold_masks.append(test_masks)
kfold_prob.append(test_probs)
print 'Summing results of ensemble'
#
res_masks = np.ndarray((dlen, 1, img_rows, img_cols), dtype=np.float32)
res_probs = np.ndarray((dlen, ), dtype=np.float32)
for i in xrange(dlen):
masks = np.ndarray((kfold, 1, img_rows, img_cols), dtype=np.float32)
probs = np.ndarray((kfold, ), dtype=np.float32)
for k in xrange(kfold):
masks[k] = kfold_masks[k][i]
probs[k] = kfold_prob[k][i]
res_masks[i] = np.mean(masks, 0)
res_probs[i] = np.mean(probs)
print('Saving ')
np.save(Learner.test_mask_res, res_masks)
np.save(Learner.test_mask_exist_res, res_probs)
"""
print timeit.timeit(code, number=100000)
'''
#def training_testing(inputlines,type):
# type 1 is splitting the data into training and testing.
fopen = open("./train.csv","r")
datalines = fopen.readlines()
#linenumbers = range(1,len(datalines)) # skipping the first line
datalines = datalines[1:]
#print linenumbers
shuffle(datalines)
percentage = int(0.8*len(datalines))
training_lines = datalines[:percentage]
print "training lines: ",len(training_lines)
[worddict,gtruth,idf_dict ] = train.preprocess(training_lines)
train.multinomial_training(worddict,gtruth,idf_dict,training_lines,"test_model_param.p")
# need to parse the testing_lines to remove the ground truth and repack as list of string.
gold_data = [int(line[0]) for line in datalines[percentage+1:]]
testing_lines =[line[2:] for line in datalines[percentage+1:] ]
print "testing lines: ",len(testing_lines)
test.multinomial_testing("test_model_param.p",testing_lines,"test_results.txt")
lista = open("test_results.txt","r").readlines()
listb = [int(element.strip().split(",")[1]) for element in lista[1:]]
comparision = map(operator.sub, gold_data, listb)
print "If score is 0 its perfect return else score is error value"
import numpy as np
from predict import load_test_data
from train import preprocess
path = '/Users/xuchenyang/Documents/third_exp/file/segnet-lr-3-32-100/'
predicted_masks = np.load( path + 'predict.npy')
imgs_test, imgs_test_mask = load_test_data()
#imgs_test_source = imgs_test.astype('float32')
imgs_test_gt = preprocess(imgs_test_mask)
predicted_masks_flat = predicted_masks.flatten()
test_gt_masks_flat = imgs_test_gt.flatten()
from sklearn import metrics
fpr, tpr, thresholds = metrics.roc_curve(test_gt_masks_flat, predicted_masks_flat, pos_label=255)
import matplotlib.pyplot as plt
#plt.plot(list(fpr),list(tpr))
plt.plot([0,1],[0,1],'k--')
line1, = plt.plot(fpr,tpr,'b',label="U-NET ROC (AUC = 0.86)")
plt.legend(handles=[line1],loc=4,prop={'size':12})
#plt.plot(list(fpr),list(tpr))
plt.xlim(0,1.0)
plt.ylim(0,1.0)
plt.xlabel("False Positive Rate (1-Specificity)")
plt.ylabel("True Positive Rate (Sensitivity)")
plt.grid()
plt.savefig(path+'roc')
import pandas as pd
from train import preprocess
from keras.models import load_model
from math import isnan
dataframe = pd.read_csv("data/test.csv")
p_ids = dataframe["PassengerId"]
data = preprocess(dataframe)
model = load_model('titanic.h5')
prediction = model.predict(data)
with open('result.csv', 'w') as file:
file.write("PassengerId,Survived\n")
for p_id, pred in zip(p_ids, prediction):
rounded_pred = int(round(pred[0]))
file.write(str(p_id) + "," + str(rounded_pred) + "\n")
run_id = version.run_id
model = mlflow.sklearn.load_model(f"models:/{model_name}/{model_stage}")
# Load data
data_path = Path(__file__).parents[0].resolve() / "data"
train_df = pd.read_csv(data_path / "train.csv")
test_df = pd.read_csv(data_path / "test.csv")
# preprocessing
impute_strats = {
"Age": round(train_df["Age"].mean()),
"Embarked": train_df["Embarked"].mode()[0],
"Cabin": "Unknown",
"Fare": round(train_df["Fare"].mean()),
}
test_df = preprocess(test_df, feature_labels, impute_strats)
test_df["Survived"] = model.predict(test_df[feature_labels])
test_df[["PassengerId", "Survived"]].to_csv(data_path / "submission.csv",
index=False)
# submit to kaggle
kaggle.api.competition_submit(
file_name=str(data_path / "submission.csv"),
message="Testing submission api",
competition="titanic",
)
time.sleep(30)
test_accuracy = kaggle.api.process_response(
kaggle.api.competitions_submissions_list_with_http_info(
"titanic"))[0]["publicScore"]
mlflow_client.log_metric(run_id, "test_accuracy", float(test_accuracy))
def test_preprocess(self):
x_train, y_train, x_dev, y_dev = preprocess()
print("x_train, y_train, x_dev, y_dev: {0},{1}".format(len(x_train), len(x_dev)))
def run_test():
BS = 128
print('Loading and preprocessing test data...')
mean, std = Learner.load_meanstd()
imgs_test, img_test_mask_gt = load_test_data()
test_img_id = load_test_ids()
imgs_test = preprocess(imgs_test)
img_test_mask_gt = preprocess(img_test_mask_gt)
imgs_test = imgs_test.astype('float32')
imgs_test -= mean
imgs_test /= std
img_test_mask_gt = img_test_mask_gt.astype('float32')
# mask_array = np.array(mask_array, dtype=np.float32)
img_test_mask_gt /= 255.0
print('Loading saved weights...')
model = get_unet(Adam(0.001))
print ('Loading weights from %s' % Learner.best_weight_path)
model.load_weights(Learner.best_weight_path)
print ('Augment')
alen, dlen = len(transforms), len(imgs_test)
test_x = np.ndarray((alen, dlen, 1, img_rows, img_cols), dtype=np.float32)
for i in range(dlen):
for j, transform in enumerate(transforms):
test_x[j,i] = transform['do'](imgs_test[i].copy())
#
print('Predicting masks on test data...')
outputs = []
asis_res = model.predict(imgs_test, batch_size=BS, verbose=1)
outputs.append(asis_res)
for j, transform in enumerate(transforms):
t_y = model.predict(test_x[j], batch_size=BS, verbose=1)
outputs.append(t_y)
#
print('Analyzing')
test_masks = np.ndarray((dlen, 1, img_rows, img_cols), dtype=np.float32)
test_probs = np.ndarray((dlen, ), dtype=np.float32)
for i in range(dlen):
masks = np.ndarray((alen+1, 1, img_rows, img_cols), dtype=np.float32)
probs = np.ndarray((alen+1, ), dtype=np.float32)
for j, t_y in enumerate(outputs):
mask, prob = t_y[0][i], t_y[1][i]
if j:
mask = transforms[j-1]['undo'](mask)
masks[j] = mask
probs[j] = prob
#
test_masks[i] = np.mean(masks, 0)
test_probs[i] = np.mean(probs)
# test_eval = model.evaluate(imgs_test, img_test_mask_gt, batch_size=BS)
print(img_test_mask_gt.shape)
print(test_masks.shape)
# pred_dir = 'preds'
# if not os.path.exists(pred_dir):
# os.mkdir(pred_dir)
# for image, image_id in zip(test_masks, test_img_id):
# # image = (image[:, :, 0] * 255.).astype(np.uint8)
# image = (image[0, :, :] * 255.)
# print(image)
# print(image.shape)
# imsave(os.path.join(pred_dir, str(image_id) + '_pred.png'), image)
print('Saving ')
np.save(Learner.test_mask_res, test_masks)
np.save(Learner.test_mask_exist_res, test_probs)
np.save(Learner.test_mask_gt, img_test_mask_gt)
[1, 0, 0, 0, 0, 0, 0, 0, 0, 0], # 0
[0, 0, 0, 0, 0, 0, 0, 0, 1, 0], # 8
[0, 0, 1, 0, 0, 0, 0, 0, 0, 0], # 2
[0, 0, 0, 0, 1, 0, 0, 0, 0, 0], # 4
]
# test data
test_digits_path = crop_image(
os.path.join(os.path.curdir, 'captcha', 'test_set',
'captcha_test.jpg'), 'test')
test_feature_set = []
for digit_path in test_digits_path:
test_feature_set.append(feature_extract(digit_path))
test_label_set = [
[0, 0, 0, 0, 0, 0, 0, 1, 0, 0], # 7
[0, 0, 0, 1, 0, 0, 0, 0, 0, 0], # 3
[1, 0, 0, 0, 0, 0, 0, 0, 0, 0], # 0
[0, 0, 0, 0, 1, 0, 0, 0, 0, 0], # 4
]
# train & predict
scaler = preprocess(train_feature_set)
scaler.transform(train_feature_set)
scaler.transform(test_feature_set)
clf = train(train_feature_set, train_label_set)
prediction = clf.predict(test_feature_set)
print(prediction)
print(np.argmax(prediction, axis=1))
print("\nEvaluation:")
dev_step(x_dev, y_dev, writer=dev_summary_writer)
print("")
path = saver.save(sess,
checkpoint_prefix,
global_step=current_step)
print("Saved model checkpoint to {}\n".format(path))
def get_args():
parser = argparse.ArgumentParser()
parser.add_argument("--prune", dest="prune", action="store_true")
parser.add_argument("--make_graph", dest="makegraph", action="store_true")
args = parser.parse_args()
return args
if __name__ == '__main__':
args = get_args()
if args.prune:
prune_filter_weight("./runs/1557076524/checkpoints/model-21300.meta",
'./runs/1557076524/checkpoints/', 128)
#filters = prune_filter_weight("./pruned_by_l2_norm/1557071581/checkpoints/model-2600.meta", './pruned_by_l2_norm/1557071581/checkpoints',90)
elif args.makegraph:
x_train, y_train, vocab_processor, x_dev, y_dev, x_test, y_test = train.preprocess(
)
make_graph(x_train, y_train, vocab_processor, x_dev, y_dev)
def predict():
model = get_unet()
# print (model.metrics_names)
# imgs_train, imgs_mask_train = load_train_data()
path_to_save_results = path + "UNET_PREDICTIONS/"
# imgs_train = preprocess(imgs_train)
# imgs_mask_train = preprocess(imgs_mask_train)
#
# # mean= np.mean(img)
# # std = np.std(imgs_mask_train)
imgs_test, imgs_test_mask = load_test_data()
mean = np.mean(imgs_test)
std = np.std(imgs_test)
# print(std)
imgs_test = preprocess(imgs_test)
imgs_test_mask = preprocess(imgs_test_mask)
imgs_test_source = imgs_test.astype('float32')
imgs_test_source -= mean
imgs_test_source /= std
imgs_test_mask = imgs_test_mask.astype('float32')
imgs_test_mask /= 255. # scale masks to [0, 1]
print('Loading saved weights...')
print('-' * 30)
model.load_weights(path + 'unet.hdf5')
print('Predicting masks on test data...')
print('-' * 30)
imgs_mask_predict = model.predict(imgs_test_source, verbose=1)
res = model.evaluate(imgs_test_source,
imgs_test_mask,
batch_size=32,
verbose=1)
res_loss = np.array(res)
np.save(path + 'predict.npy', imgs_mask_predict)
np.savetxt(path + 'res_loss.txt', res_loss)
predicted_masks = np.load(path + 'predict.npy')
predicted_masks *= 255
imgs_test, imgs_test_mask = load_test_data()
for i in range(imgs_test.shape[0]):
img = resize(imgs_test[i], (96, 96), preserve_range=True)
img_mask = resize(imgs_test_mask[i], (96, 96), preserve_range=True)
im_test_source = Image.fromarray(img.astype(np.uint8))
im_test_masks = Image.fromarray((img_mask.squeeze()).astype(np.uint8))
im_test_predict = Image.fromarray(
(predicted_masks[i].squeeze()).astype(np.uint8))
im_test_source_name = "Test_Image_" + str(i + 1) + ".png"
im_test_predict_name = "Test_Image_" + str(i + 1) + "_Predict.png"
im_test_gt_mask_name = "Test_Image_" + str(i + 1) + "_OriginalMask.png"
im_test_source.save(
os.path.join(path_to_save_results, im_test_source_name))
im_test_predict.save(
os.path.join(path_to_save_results, im_test_predict_name))
im_test_masks.save(
os.path.join(path_to_save_results, im_test_gt_mask_name))
message = "Successfully Saved Results to " + path_to_save_results
print message
import numpy as np
import pandas as pd
import pickle
from train import preprocess
outfile = 'model.sav'
# Parsing script arguments
parser = argparse.ArgumentParser(description='Process input')
parser.add_argument('tsv_path', type=str, help='tsv file path')
args = parser.parse_args()
# Reading input TSV
data = pd.read_csv(args.tsv_path, sep="\t")
ids = data['id'].copy()
X, y_true = preprocess(data, is_train=False)
#####
# TODO - your prediction code here
model = pickle.load(open(outfile, 'rb'))
log_pred = model.predict(X)
y_pred = np.exp(log_pred)
# Example:
prediction_df = pd.DataFrame(columns=['id', 'revenue'])
prediction_df['id'] = ids
prediction_df['revenue'] = y_pred
####
# TODO - How to export prediction results
prediction_df.to_csv("prediction.csv", index=False, header=False)
import numpy as np
from train import get_unet, preprocess
from data import load_test_data, load_train_data
imgs_train, imgs_mask_train = load_train_data()
imgs_train = preprocess(imgs_train)
imgs_train = imgs_train.astype('float32')
mean = np.mean(imgs_train)
std = np.std(imgs_train)
imgs_test, imgs_id_test = load_test_data()
imgs_test = preprocess(imgs_test)
imgs_test = imgs_test.astype('float32')
imgs_test -= mean
imgs_test /= std
model = get_unet()
model.load_weights("final.h5")
imgs_mask_test = model.predict(imgs_test, verbose=1)
np.save('imgs_mask_test_final.npy', imgs_mask_test)
"""
print timeit.timeit(code, number=100000)
'''
#def training_testing(inputlines,type):
# type 1 is splitting the data into training and testing.
fopen = open("./train.csv", "r")
datalines = fopen.readlines()
#linenumbers = range(1,len(datalines)) # skipping the first line
datalines = datalines[1:]
#print linenumbers
shuffle(datalines)
percentage = int(0.8 * len(datalines))
training_lines = datalines[:percentage]
print "training lines: ", len(training_lines)
[worddict, gtruth, idf_dict] = train.preprocess(training_lines)
train.multinomial_training(worddict, gtruth, idf_dict, training_lines,
"test_model_param.p")
# need to parse the testing_lines to remove the ground truth and repack as list of string.
gold_data = [int(line[0]) for line in datalines[percentage + 1:]]
testing_lines = [line[2:] for line in datalines[percentage + 1:]]
print "testing lines: ", len(testing_lines)
test.multinomial_testing("test_model_param.p", testing_lines,
"test_results.txt")
lista = open("test_results.txt", "r").readlines()
listb = [int(element.strip().split(",")[1]) for element in lista[1:]]
comparision = map(operator.sub, gold_data, listb)
print "If score is 0 its perfect return else score is error value"
