def Average(lst):
lst = np.array(lst)
print(lst.shape)
for l in lst:
if l is None:
logger.info("None found")
exit(1)
return sum(lst) / len(lst)
def save_test_features(data):
start_timestamp = time.time()
dataset = {}
logger.info("Entered on reduce test features process")
for user, paths in data.items():
features = [extract_features_of_roi(path)[0] for path in paths]
dataset.update({user: [features, paths]})
# saving test database
with open(TEST_PATH, "wb") as file:
pickle.dump(dataset, file)
logger.info("Extraction concluded in {} seconds.".format(time.time() -
start_timestamp))
def get_features_from_user(name):
# loading train a catalogues
with open(TRAIN_PATH, "rb") as file:
train = pickle.load(file)
# separing train and test features
# logger.info('Getting features from user {}'.format(name))
if name in train:
t = np.array(train[name][0])
return (np.reshape(t, (t.shape[0], t.shape[2])))
else:
logger.info("User {} not enrolled on dataset".format(name))
return None, None, None
def save_train_features(data):
start_timestamp = time.time()
dataset = {}
full_dataset = {}
logger.info("Entered on reduce train features process")
for user, paths in data.items():
features = [extract_features_of_roi(path)[0] for path in paths]
avg = np.array(Average(features))
arg = np.array(l2(vector - avg) for vector in features).argmin()
dataset.update({user: avg})
full_dataset.update({user: [features, paths]})
logger.info("Extraction concluded in {} seconds.".format(time.time() -
start_timestamp))
with open(DATAKEYS_PATH, "wb") as file:
pickle.dump(dataset, file)
with open(TRAIN_PATH, "wb") as file:
pickle.dump(full_dataset, file)
return 0
def avaliate(classes, ctrs, dataset, n):
sum = 0
qtd = 0
total_wrong = 0
for user, fts in dataset.items():
true_centroid = [idc for idc, c in enumerate(classes) if user in c]
report = {}
for i in range(len(fts[0])):
winners = compute_distance(ctrs, fts[0][i], n)
st = "centers_"
st2 = ''
for i in range(len(winners)):
st2 += str(winners[i]) + "_"
st += st2
if st not in report.keys():
report.update({st: []})
distances = []
for i in range(len(winners)):
distances.append(l2(ctrs[true_centroid] - ctrs[winners[i]]))
# distance = l2(ctrs[true_centroid] - ctrs[winner])
minimum = min(distances)
if minimum > 0:
total_wrong += 1
report[st].append(minimum)
sum += minimum
qtd += 1
logger.info("Report for user {}: {}".format(user, report))
logger.info("Quantization error: {}".format(sum / qtd))
logger.info("Miss prediction tax: {}".format(total_wrong / qtd))
def cut_clusters(classes, centers, thrs=7):
centers = np.array(centers)
classes = np.array(classes)
to_cut = {}
for idc, c in enumerate(classes):
if (len(c) <= thrs):
to_cut.update({idc: c})
index = [index for index, users in to_cut.items()]
new_classes = [classes[i] for i in range(len(classes)) if i not in index]
new_centers = [centers[i] for i in range(len(centers)) if i not in index]
logger.info("Index of wrong clusters {}".format(index))
for idc, c in to_cut.items():
logger.info("Cluster {} has minus than {} users".format(idc, thrs))
# logger.info("Classes vector before remotion: {}".format(classes))
# logger.info("Classes vector after remotion: {}".format(classes))
for name in c:
vets = get_features_from_user(name)
new_cluster = reallocate_user(vets, new_centers)
logger.info("New cluster for user {} is cluster number {}".format(
name, new_cluster))
new_classes[new_cluster].append(name)
new_classes = np.array(new_classes)
new_centers = np.array(new_centers)
print("New shape for classes and centers: {}, {}".format(
new_classes.shape, new_centers.shape))
return new_classes, new_centers