def load_data():
vec = DictVectorizer()
file_pairs = []
for i in range(2014, 2019):
file_pairs.append([('data/players_' + str(i) + '.csv'), ('data/all_stars_' + str(i) + '.txt'), i])
var, labels, names = get_data(file_pairs)
vec = vec.fit_transform(var).toarray()
labels = np.array(labels)
vec = np.array(vec)
return vec, labels, names
def init_players():
players = import_data.get_data("KRates.csv")
lg_avg = set_lg_avg_stats(players)
id = "andeg001"
set_avg(players[id].stats["SO"])
set_w_avg(players[id].stats["SO"])
set_w_var(players[id].stats["SO"])
set_uncertainty(players[id].stats["SO"])
calc_projection(players[id].stats["SO"],lg_avg)
print_results(players[id],lg_avg)
def init_players():
players = import_data.get_data("KRates.csv")
lg_avg = set_lg_avg_stats(players)
id = "andeg001"
set_avg(players[id].stats["SO"])
set_w_avg(players[id].stats["SO"])
set_w_var(players[id].stats["SO"])
set_uncertainty(players[id].stats["SO"])
calc_projection(players[id].stats["SO"], lg_avg)
print_results(players[id], lg_avg)
def main(argv):
initial_dataset = import_data.get_data()
processing_dataset = import_data.get_data()
processing_dataset = centroids.randomize_centroids(processing_dataset, number_of_centroids)
verification_results = do_clustering(processing_dataset)
while True:
if is_clustering_successful(verification_results):
output_success_rates(processing_dataset, initial_dataset)
break
elif is_too_many_attempts():
print("TOO MANY ATTEMPTS")
break
else:
centroids_ids = list(map(lambda x: x["id"], verification_results))
processing_dataset = centroids.setup_centroids_new(processing_dataset, verification_results)
verification_results = do_clustering(processing_dataset)
print("FINAL, ITERATIONS COUNT: ", iteration)
print("TIME DURATION %s seconds" % (time.time() - start_time))
'''
import matplotlib.pyplot as plt
import import_data as imp_data
import plot_trap_site as plt_trap
import config
import sys
import numpy as np
import temperature_single_atom as sim
'''
Here get Boolean data.
Sequence of zeros and ones
corresponding to trap site occupied or not. That simple.
'''
x_values, initial_bool, assembled_bool, assembled_list_bool, recapture_bool, atom_loss_bool = imp_data.get_data(
)
# Plot Histogram with number of atoms in each run
plt_trap.atom_number(initial_bool)
'''
Plot occupation probability for each trap site.
'''
plt_trap.init_plot(initial_bool)
calc_diff = False
if calc_diff == True:
# If True will plot image on 'assembled bool - initial'
plt_trap.diff_plot(initial_bool, assembled_bool)
# Filter data for rearrangement ??
filter = False
# This file creates and compuets the norms for all the layers present in a model # Import the uwimg library from uwimg import * from model_definition import softmax_model from import_data import get_data # FIRST LOAD THE DATA train_data = get_data()[0] test_data = get_data()[1] # Now iterate through the model layers def iterate_over(model, iters, batchsize): # Create a model for iter in range(iters): calculate_G1(Model, data.X, data) print_matrix(Model.G1) psi = create_psi(data.X, data.y, 0.1) calculate_G2(Model, train.X, data, psi) # Now calculate the running average of Psi for all the layers in the model norm1 = 0 for layers in Model.n: psi = running_average(psi, (Model.layers+iter).G1, (Model.layers+iter).G2) norm1 = calculate_delta_norm(norm1 ,psi) if __name__ == "__main__": batchsize = 128 data = random_batch(test_data ,batchsize) Model = softmax_model(data.X.cols, data.y.cols)
nuc_list.append(nuc)
cyto_list.append(cyto)
nuc_cyto_list.append(cv2.bitwise_or(nuc, cyto))
return nuc_list, cyto_list, nuc_cyto_list, names, discard
def apply_mask(img_gray_list, mask_list):
out = []
for idx, each in enumerate(img_gray_list):
out.append(cv2.bitwise_and(each, mask_list[idx]))
return out
segmented_list = import_data.get_data(PATH_TO_EXPERT_HERLEV_SEGMENTATION)
cell_image_list = import_data.get_data(PATH_TO_HERLEV_IMGS)
images_list, name_and_label_imgs_og = zip(*cell_image_list)
segmented_masks, name_and_label_exp_segment = zip(*segmented_list)
cv2.imshow(name_and_label_exp_segment[EVAL_INDEX], segmented_masks[EVAL_INDEX])
cv2.imshow(name_and_label_imgs_og[EVAL_INDEX], images_list[EVAL_INDEX])
# cv2.waitKey()
b_list_seg, g_list_seg, r_list_seg = separate_into_channels(segmented_masks)
compare = np.hstack(
[b_list_seg[EVAL_INDEX], g_list_seg[EVAL_INDEX], r_list_seg[EVAL_INDEX]])
cv2.imshow('check', compare)
gray_list = rgb_to_gray(images_list)