def main(*args):
try:
twitter_user, tweet_id, flag_dict = gr(args[0])
except Exception as e:
print('No args passed to pipeline.py main()')
print(e)
twitter_user, tweet_id, flag_dict = gr()
prep(tweet_id, flag_dict)
dp = data_paths(tweet_id=tweet_id, twitter_user_id=twitter_user)
tweet_html = ge(tweet_id)
print('\nUploading to Google Big Query...')
if flag_dict['big_query'] != 'off':
up_gbq(tweet_id)
print('Done.')
async def on_message(message):
global topic, lan, found, comment_flag
#traduction message user
if message.content not in ['/help', '/h']:
ms, lan = preproc.transin(message.content)
msg = preproc.prep(ms)
else:
msg = message.content
if message.author != client.user:
#Gestion de la conversation
if msg in library.lib and comment_flag == False:
if msg in ['hello'] or message.content in ['/help', '/h']:
await message.channel.send(file=discord.File(
'/home/roger/anaconda3/projetIA/Chat_bot/ChatBot/robot.png'
))
res = ct.chatter_bot_conv(msg)
#reponse dans la langue user
resp = preproc.transout(res, lan)
await message.channel.send(resp)
else:
#Gestion de la recherche en bdd
try:
if msg not in library.lib and preproc.transout(
msg, 'en'
) != 'no' and comment_flag == False and preproc.transout(
msg, 'en') != 'yes':
topic = topic_clf.predict([msg])[0]
print("search", topic_clf.predict([msg])[0])
resp = search.find_question_answer(msg, topic)
found = True
resp = preproc.transout(resp, lan)
await message.channel.send(resp[:2000])
except NameError:
if preproc.transout(msg,
'en') != 'no' and comment_flag == False:
response = preproc.transout(
'Pouvez-vous m\'en dire plus ?', lan)
await message.channel.send(response)
# fin de conversation
if found == True and msg not in library.lib and preproc.transout(
msg, 'en') != 'no' and preproc.transout(
msg, 'en') != 'yes' and comment_flag == False:
resp = preproc.transout(library.end_of_conv[0],
lan) + library.end_of_conv[1]
found = False
await message.channel.send(resp)
if preproc.transout(msg, 'en') == 'no' and comment_flag == False:
resp = preproc.transout(library.bot_end_conv[0], lan)
await message.channel.send(resp)
comment_flag = True
if preproc.transout(msg, 'en') != 'no' and comment_flag == True:
feed_back.add_feeback({
"message": message.content,
"date": str(datetime.date.today())
})
comment_flag = False
def run():
df_pre = prep()
df_sampled = sampler(df_pre)
# df_reduced_dims = reduction(df_sampled)
# quick_grapher.graph(df_reduced_dims)
# Model fitting with SVM
#df_B = svc(df_sampled)
df_C = KNN(df_sampled)
def main():
rw = readwrite()
trainDataset = rw.readcsv(trainFilePath)
prepros_object = prep(test_first_level_split, test_second_level_split,
trainDataset)
(first_train, first_test, second_train,
second_test) = prepros_object.labelPaths(Routine_k)
#each of the above four are comprised of a triple: (data, path, routine/non-routine)
#Build the model and store it on disk
topLevelClassifier(first_train, first_test, DirToDumpModelsTop)
secondLevelClassifier(second_train, second_test, DirToDumpModelsSecond)
def run(): df_pre = prep() #df_B = svc(df_pre) #df_C = KNN(df_pre) df_D = NN(df_pre)