def main():
#url = input("Enter the url of blog: ")
app = UI.App()
url = str(app.returnURL())
print("URL = " + str(url) + "\n------")
posts = ws.scrapeBlog(url)
print("Scrapingok \n------")
for post in posts:
print("into the loop")
filename = ws.scrapePost(post)
#path = input("Enter the name of the file: ")
#The League of Extraordinary Gentlemen – Akaash Preetham
text = dr.readDoc(filename)
#text = input("Enter text here at main: ")
#sortedWords = wc.countWords(text)
#for word in sortedWords:
# print(str(word[0]) + '-' + str(word[1]))
#wordcount = wc.countWords(text)
#dv.visualizeData(wordcount)
dv.visualizeData(text, filename)
def tweet_picture_week():
week_visualization()
with open("grow.png", "rb") as imagefile:
imagedata = imagefile.read()
t_upload = Twitter(domain='upload.twitter.com',
auth=OAuth(
token=getenv("twitter_token"),
token_secret=getenv("token_secret"),
consumer_key=getenv("twitter_consumer_key"),
consumer_secret=getenv("twitter_consumer_secrets")))
id_img1 = t_upload.media.upload(media=imagedata)["media_id_string"]
t.statuses.update(status=DataV.text_week(), media_ids=",".join([id_img1]))
import DataVisualization
#Initialize data
years = [
1980, 1981, 1982, 1983, 1984, 1985, 1986, 1987, 1988, 1989, 1990, 1991,
1992, 1993
]
population = [
5000, 5100, 5800, 6800, 6600, 6700, 6900, 7000, 7500, 8000, 8800, 10000,
10982, 12000
]
#Draw line chart
dataVisualization = DataVisualization.DataVisualization(
years, population, "Population", "Population", "Year")
dataVisualization.DrawLineChart()
interpolated_x2 = np.arange(0, pmobj.get_end_time(), 0.032)
interpolated_y2 = f(interpolated_x2)
plt.plot(interpolated_x2 , interpolated_y2)
'''
plt.show()
exit()
piano_roll = pmobj.get_piano_roll(fs=100)
print(piano_roll.shape)
dv.showPiano(piano_roll)
piano_roll = get_piano_roll_mod(ins, fs=100)
print(piano_roll.shape)
dv.showPiano(piano_roll)
print('Processing File Pair %d/%d Name:%s' % (index + 1, N, pair[1]))
#(inputnp, times) = prepFunc.procWaveData(pair[1])
#print("Input shape:{0}".format(inputnp.shape))
#origin_bins = math.ceil(pmobj.get_end_time() * prepFunc.sr / 512)
#mytimes = np.arange(0,origin_bins,dtype="float")*(0.032)
output = model.forward(inputTensor)
print("Input shape:{0}".format(inputTensor.shape))
print("Ground-truth shape:{0}".format(labelTensor.shape))
print("Predict-label shape:{0}".format(output.shape))
'''
if usingThreshold:
output = output > 0.5
fscore = skm.f1_score(labelTensor.cpu().numpy(), output.cpu().detach().numpy(), average='samples')
print(fscore)
fscorelist.append(fscore)
'''
if usingThreshold:
output = output > 0.5
dv.showSongData2(inputNP, labelTensor, output)
'''
if pairIndex % plotEverySong == 0 and pairIndex > 0:
plt.plot(losslist)
plt.savefig(lossFiguresPath+'result_%d.png' % pairIndex)
'''
totalElapsedTime = time.time() - startTime
print('Total training time : ' +
time.strftime('%H:%M:%S', time.gmtime(totalElapsedTime)))
plt.plot(fscorelist)
plt.show()
totalElapsedTime = time.time() - startTime
#dv.showSongData(inputNP, labelTensor, output)
outputDNN = outputDNN > 0.5
fscoreDNN = skm.f1_score(labelTensor.cpu().numpy(),
outputDNN.cpu().detach().numpy(),
average='samples')
print("DNN Fscore %f.6" % fscoreDNN)
outputRNN = outputRNN > 0.4
fscoreRNN = skm.f1_score(labelTensor.cpu().numpy(),
outputRNN.cpu().detach().numpy(),
average='samples')
print("RNN Fscore %f.6" % fscoreRNN)
dv.showSongDataWithModels(inputNP, labelTensor, [outputDNN, outputRNN])
'''
if pairIndex % plotEverySong == 0 and pairIndex > 0:
plt.plot(losslist)
plt.savefig(lossFiguresPath+'result_%d.png' % pairIndex)
'''
totalElapsedTime = time.time() - startTime
print('Total time : ' +
time.strftime('%H:%M:%S', time.gmtime(totalElapsedTime)))
break
totalElapsedTime = time.time() - startTime
print('Total time : ' +
print("number of results:", str(len(framework.results)))
# Export results to a pandas panel
print("exporting results.")
results = framework.export_results_as_panel()
print(results.to_frame(filter_observations=False))
# Data Visualization
import matplotlib.pyplot as plt
import math
# Generate performance profiles for all solver configurations
# Graph performance profile:
plt.figure()
dv.plot_performance_profile(results)
plt.draw()
# Graph time vs. big(small)^2
plt.figure()
dv.plot_scatter_by_config(results, "max_big_small_squared", "solve_time")
plt.draw()
# Graph time vs. number of scalar variables
plt.figure()
dv.plot_scatter_by_config(results,
"num_scalar_variables",
"solve_time",
logx=True,
logy=True)
plt.draw()
try:
print('Директория удалена')
shutil.rmtree(img_dir)
except:
print('Директория не найдена')
# создание директории для сохранения изображений
os.makedirs(img_dir, exist_ok=True)
# создается экземпляр класса обработки кадров
# задаются параметры - число строк, число столбцов
fp = FramesProcessing.FramesProcess(3, 15)
# экземпляр класса для построения и отображения графиков
vd = DataVisualization.VisualData()
vd.set_config(300, 550, [0, 0, 0]) # параметры изображения графиков
histo = vd.set_bgimage() # "холст" для рисования графиков
# ======================================================================
limit_ignore = 20 # число игнорируемых кадров
ignore_count = 0 # счетчик игнорируемых кадров
cnt_image = 0 # счетчик сделаных снимков
write_ok = False # флаг разрешения записи файла
framescnt = 0 # счетчик кадров
old_array = [0] # предыдущий массив
new_array = [0] # новый массив
cap = cv2.VideoCapture(-1)
cv2.namedWindow('histo')
CNN_output[i] = CNN_model.forward(CNN_inputTensor[i].unsqueeze(0))
'''
# loss = model.Loss(output, labelTensor)
if useThreshold:
# CNN_output = CNN_output > 0.5
RNN_output = RNN_output > 0.5
# DNN_output = DNN_output > 0.5
# print("Input shape:{0}".format(inputTensor.shape))
print("Ground-truth shape:{0}".format(labelTensor.shape))
#print("Predict-label shape:{0}".format(output.shape))
print(RNN_output[38032])
exit()
dv.showSongData3(inputNP, labelTensor, RNN_output)
# DNN_FScore.append(f1_score(labelTensor.detach().cpu().numpy(), DNN_output.detach().cpu().numpy(),average ='samples'))
# CNN_FScore.append(f1_score(labelTensor.detach().cpu().numpy(), CNN_output.detach().cpu().numpy(),average ='samples'))
# RNN_FScore.append(f1_score(labelTensor.detach().cpu().numpy(), RNN_output.detach().cpu().numpy(),average ='samples'))
totalElapsedTime = time.time() - startTime
print('Total training time : ' + time.strftime('%H:%M:%S', time.gmtime(totalElapsedTime)))
# np.save("CNN_final", CNN_output.detach().cpu().numpy())
# np.save("RNN_final", RNN_output.detach().cpu().numpy())
# np.save("DNN_final", DNN_output.detach().cpu().numpy())
# print(np.mean(DNN_FScore))
print(np.mean(CNN_FScore))
# print(np.mean(RNN_FScore))