def create_word_cloud():
""" create the cloud of word with wordcloud_fa module"""
# mask.jpg is a image in black and white picture that word will write in that
mask_array = np.array(Image.open("mask.jpg"))
wc = WordCloudFa(persian_normalize=True,
include_numbers=False,
mask=mask_array,
background_color="white",
collocations=False)
with open('telegramtxt.txt', 'r') as file:
text = file.read()
frequencies = wc.process_text(text)
avr = 0
count = 0
frequencies = {k: v for k, v in frequencies.items() if v > 1}
for k, v in frequencies.items():
count += 1
avr += v
avr = avr // count
print(f'avr of word count : {avr}')
frequencies = {k: v for k, v in frequencies.items() if v > avr}
frequencies = {
k: v
for k, v in sorted(
frequencies.items(), key=lambda item: item[1], reverse=True)
}
word_cloud = wc.generate_from_frequencies(frequencies)
image = word_cloud.to_image()
image.save('cloud.png')
def save_word_cloud(user_name: str, api):
raw_tweets = []
for tweet in tweepy.Cursor(api.user_timeline, id=user_name).items():
raw_tweets.append(tweet.text)
# Normalize words
tokenizer = WordTokenizer()
lemmatizer = Lemmatizer()
normalizer = Normalizer()
stopwords = set(
list(
map(lambda w: w.strip(),
codecs.open(STOPWORDS_PATH, encoding='utf8'))))
words = []
for raw_tweet in raw_tweets:
raw_tweet = re.sub(r"[,.;:?!،()]+", " ", raw_tweet)
raw_tweet = re.sub('[^\u0600-\u06FF]+', " ", raw_tweet)
raw_tweet = re.sub(r'[\u200c\s]*\s[\s\u200c]*', " ", raw_tweet)
raw_tweet = re.sub(r'[\u200c]+', " ", raw_tweet)
raw_tweet = re.sub(r'[\n]+', " ", raw_tweet)
raw_tweet = re.sub(r'[\t]+', " ", raw_tweet)
raw_tweet = normalizer.normalize(raw_tweet)
raw_tweet = normalizer.character_refinement(raw_tweet)
tweet_words = tokenizer.tokenize(raw_tweet)
tweet_words = [
lemmatizer.lemmatize(tweet_word).split('#', 1)[0]
for tweet_word in tweet_words
]
tweet_words = list(filter(lambda x: x not in stopwords, tweet_words))
words.extend(tweet_words)
if len(words) == 0:
return
# Build word_cloud
mask = np.array(Image.open(MASK_PATH))
clean_string = ' '.join([str(elem) for elem in words])
clean_string = arabic_reshaper.reshape(clean_string)
clean_string = get_display(clean_string)
word_cloud = WordCloudFa(persian_normalize=False,
mask=mask,
colormap=random.sample(cmaps, 1)[0],
background_color=BACKGROUND_COLOR,
include_numbers=False,
font_path=FONT_PATH,
no_reshape=True,
max_words=1000,
min_font_size=2)
wc = word_cloud.generate(clean_string)
image = wc.to_image()
image.save(word_cloud_address)
def create_word_clod(all_cluster_hashtags, path):
'''
:param all_cluster_hashtags: something like :list of Counter() ,each Counter have hashtags with
the number of that hashtag in that cluster
:return:
'''
for i, hashtags in enumerate(all_cluster_hashtags):
wodcloud = WordCloudFa()
wc = wodcloud.generate_from_frequencies(
dict(hashtags.most_common()[0:5]))
image = wc.to_image()
# image.show()
image.save('{0}/cluster_{1}.png'.format(path, i))
def get_image(UserID, chat_id, message, bot, update):
if len(UserID) < 2:
message(chat_id=chat_id.chat_id, text="آی دی نامعتبر است!")
else:
message(chat_id=chat_id.chat_id, text="در حال اتصال به اینستاگرام...")
allword = query.start(UserID, chat_id, message, bot, update)
allword_edited = removeWeirdChars(allword)
my_wordcloud = WordCloudFa(font_path="Sahel.ttf",
background_color="white",
width=720,
height=1280,
margin=2).generate(allword_edited)
image = my_wordcloud.to_image()
saved_dir = 'analysis/' + str(UserID) + '.jpg'
image.save(saved_dir)
message(chat_id=chat_id.chat_id, text="درحال ارسال عکس...")
return saved_dir
def execute(self):
numberOfPages = 1
numberOfTweetsPerPage = 200
counter = 0
cloud = ""
txt = ""
if self.numberOfTweets > 200:
numberOfPages = ceil(self.numberOfTweets / 200)
else:
numberOfTweetsPerPage = self.numberOfTweets
for i in range(numberOfPages):
tweets = self.api.user_timeline(screen_name=self.username,
count=numberOfTweetsPerPage,
page=i)
for each in tweets:
cloud = each.text
cloud = re.sub(r'[A-Za-z@_]*', '', cloud)
counter += 1
txt = txt + ' ' + each.text
txt = re.sub(r'[A-Za-z@]*', '', txt)
twitter_mask = np.array(
Image.open(path.join(self.d, "templates/cloud/twitter-logo.jpg")))
stop = [
'می', 'من', 'که', 'به', 'رو', 'از', 'ولی', 'با', 'یه', 'این',
'نمی', 'هم', 'شد', 'ها', 'اما', 'تو', 'واقعا', 'در', 'نه', 'دارم',
'باید', 'آره', 'برای', 'تا', 'چه', 'کنم', 'بود', 'همه', 'دیگه',
'ای', 'اون', 'تی', 'حالا', 'بی', 'د', 'چرا', 'بابا', 'منم', 'کیه',
'توی', 'نیست', 'چی', 'باشه', 'که', 'بودم', 'می کنم', 'که', 'اینه',
'بهتر', 'داره', 'اینه', 'که', 'کردن', 'می', 'کن', 'بعد', 'دیگه',
'', '', '', ''
]
wc = WordCloudFa(
# font_path='IranNastaliq.ttf',
persian_normalize=True,
max_words=1000,
margin=0,
width=3000,
height=2500,
min_font_size=1,
max_font_size=1000,
background_color=self.backGround,
mask=twitter_mask,
include_numbers=False,
collocations=False)
wc.add_stop_words(stop)
wc.generate(txt)
directory = 'static/images/' + self.username + '.png'
directory = path.join(self.d, directory)
image = wc.to_image()
image.save(directory)
def show_chat_word_cloud(directory):
with codecs.open(os.path.join(directory, 'chats.txt'),
'r',
encoding='utf8') as file:
print("Start putting words in picture")
mask_array = np.array(Image.open("telegram.png"))
wordcloud = WordCloudFa(persian_normalize=True,
mask=mask_array,
collocations=False)
stop_words = []
with codecs.open("stop_words.txt", 'r', encoding='utf8') as words:
for word in words:
stop_words.append(word[:-2])
wordcloud.add_stop_words(stop_words)
text = delete_extra_characters(file.read())
wc = wordcloud.generate(text)
image = wc.to_image()
image.show()
image.save(os.path.join(directory, 'wordcloud.png'))
def main():
database = DB("ganjoor.s3db")
database.connect()
rows = database.select("""select p.cat_id,v.text
from poem as p join verse as v on p.id=v.poem_id
where p.cat_id=24
""")
f = open("verses.txt", "w")
for r in rows:
f.write(r[1])
f.write('\n')
f.close()
wc = WordCloudFa(width=1200, height=800)
with open('verses.txt', 'r') as file:
text = file.read()
word_cloud = wc.generate(text)
image = word_cloud.to_image()
image.show()
image.save('hafez.png')
def __init__(self,
mask=None,
size=900,
stop_words_addr=default_stop_words_path,
mask_addr=None):
self.hazm_normalizer = hazm.Normalizer()
self.parsivar_normalizer = parsivar.Normalizer()
self.stemmer = hazm.Stemmer()
self.lemmatizer = hazm.Lemmatizer()
self.stop_words = set(hazm.stopwords_list(stop_words_addr))
mask = np.array(
Image.open(mask_addr)) if mask_addr is not None else None
self.generator = WordCloud(width=size,
height=size,
include_numbers=False,
persian_normalize=False,
collocations=True,
mask=mask,
background_color='white')
from wordcloud_fa import WordCloudFa
wodcloud = WordCloudFa(no_reshape=True,
persian_normalize=True,
include_numbers=False,
collocations=False,
width=800,
height=400)
text = ""
with open('persian-example.txt', 'r') as file:
text = file.read()
wc = wodcloud.generate(text)
image = wc.to_image()
image.show()
image.save('persian-example.png')
def remove_punctuations(text):
translator = str.maketrans('', '', punctuations_list)
return text.translate(translator)
words = remove_punctuations(words)
words = re.sub(
'http[s]?://(?:[a-zA-Z]|[0-9]|[$-_@.&+]|[!*\(\),]|(?:%[0-9a-fA-F][0-9a-fA-F]))',
'',
words,
flags=re.MULTILINE)
words = re.sub(r"@(\w+)", ' ', words, flags=re.MULTILINE)
wordcloud = WordCloudFa(persian_normalize=True,
stopwords=list(STOPWORDS) +
hazm.stopwords_list(),
include_numbers=False,
background_color='white',
width=700,
height=500)
frequencies = wordcloud.process_text(words)
wc = wordcloud.generate_from_frequencies(frequencies)
image = wc.to_image()
st.image(image)
# Dataframe
st.subheader('**Data**')
st.write(data)
# Random Tweet
col1, col2 = st.beta_columns(2)
with col1:
st.markdown('')
from wordcloud_fa import WordCloudFa
wodcloud = WordCloudFa(persian_normalize=True,
include_numbers=False,
background_color="white")
text = ""
with open('mixed-example.txt', 'r') as file:
text = file.read()
wc = wodcloud.generate(text)
image = wc.to_image()
image.show()
image.save('mixed-example.png')
def make_wordcloud(font_path=None,
text_path='tweets/result.txt',
bw_img="input/southpark1.png",
img="input/southpark2.png",
add_stopwords=ADD_STOPWORDS,
bg_color='black',
include_numbers=True,
max_words=500,
random_color=False):
assert type(add_stopwords) == type(list())
# get data directory (using getcwd() is needed to support running example in generated IPython notebook)
d = path.dirname(__file__) if "__file__" in locals() else getcwd()
# load text
text = open(path.join(d, text_path), encoding='utf-8').read()
# load image. This has been modified in gimp to be brighter and have more saturation.
image = np.array(Image.open(path.join(d, bw_img)))
# subsample by factor of 3. Very lossy but for a wordcloud we don't really care.
mask_color = np.array(Image.open(path.join(d, img)))
# create mask white is "masked out"
twitter_mask = image.copy()
twitter_mask[twitter_mask.sum(axis=2) == 0] = 255
# some finesse: we enforce boundaries between colors so they get less washed out.
# For that we do some edge detection in the image
edges = np.mean([
gaussian_gradient_magnitude(mask_color[:, :, i] / 255., 2)
for i in range(3)
],
axis=0)
twitter_mask[edges > .01] = 255
# Add another stopword
stop_words = stopwords.words('english')
##stop_words_fa = stopwords.words('farsi')
for word in add_stopwords:
STOPWORDS.add(Normalizer().normalize(word))
stop_words.extend(STOPWORDS)
stop_words.extend(EN_STOPWORDS)
stop_words = set(stop_words)
# Getting rid of the stopwords
text_list = [word for word in text.split() if word not in stop_words]
# Converting the list to a text
text = ' '.join([str(elem) for elem in text_list])
text.replace('\u200c', '')
# Generate a word cloud image
wordcloud = WordCloudFa(font_path=font_path,
persian_normalize=True,
include_numbers=include_numbers,
max_words=max_words,
stopwords=stop_words,
margin=0,
width=3000,
height=3000,
min_font_size=1,
max_font_size=2300,
random_state=True,
background_color=bg_color,
mask=twitter_mask,
relative_scaling=0,
repeat=True).generate(text)
if not random_color:
image_colors = ImageColorGenerator(mask_color)
wordcloud.recolor(color_func=image_colors)
image = wordcloud.to_image()
image.show()
image.save('output/twitter_mask.png')
exit()
text = ""
print("cleaning")
text = " ".join([clean_word(word) for word in raw_text.split()])
#################################
print_stats(text)
print("generating cloud")
mask_array = np_array(Image.open(MASK))
wc_instance = WordCloudFa(
width=900,
height=900,
background_color=BG_COLOR,
font_path=FONT,
mask=mask_array,
persian_normalize=True,
include_numbers=False,
stopwords=load_stop_words(),
)
word_cloud = wc_instance.generate(text)
result_image = word_cloud.to_image()
result_image.save(RESULT_FILE_ADD)
result_image.show()
#print( "\n----\n".join(tweets_simple) )
to_print = "\n\n".join(tweets_simple)
f = open("out/cleaned.txt", "w")
f.write(to_print)
f.close()
#######################################
mask_array = np.array(Image.open("masks/tw.png"))
with open('out/cleaned.txt', 'r') as file:
text = file.read()
wc = WordCloudFa(
width=900,
height=900,
background_color="white",
font_path="fonts/font2.ttf",
mask=mask_array,
persian_normalize=True,
include_numbers=False,
)
word_cloud = wc.generate(text)
image = word_cloud.to_image()
image.save(f"out/{idish}.png")
image.show()
class WordCloudGen:
"""Word Cloud Generator"""
def __init__(self,
mask=None,
size=900,
stop_words_addr=default_stop_words_path,
mask_addr=None):
self.hazm_normalizer = hazm.Normalizer()
self.parsivar_normalizer = parsivar.Normalizer()
self.stemmer = hazm.Stemmer()
self.lemmatizer = hazm.Lemmatizer()
self.stop_words = set(hazm.stopwords_list(stop_words_addr))
mask = np.array(
Image.open(mask_addr)) if mask_addr is not None else None
self.generator = WordCloud(width=size,
height=size,
include_numbers=False,
persian_normalize=False,
collocations=True,
mask=mask,
background_color='white')
def get_word_cloud(self, msgs):
return self.generator.generate_from_text(
self._preprocess(msgs)).to_image()
def _preprocess(self, msgs):
words = []
for msg in msgs:
msg = re.sub(r"https?:\/\/\S*", "",
msg) # https://github.com/MasterScrat/Chatistics
msg = re.sub(r"\@\S*", "", msg)
msg = self._normalize(msg)
msg = msg.replace("ؤ", "و")
msg = msg.replace("أ", "ا")
msg = self._remove_punctuations(msg)
msg = self._remove_weird_chars(msg)
msg = self._remove_postfixes(msg)
for word in msg.split():
if self._is_stop_word(word):
word = ""
if word:
# word = self.stemmer.stem(word)
word = word.replace(u"\u200c", "")
words.append(word)
return " ".join(words)
def _normalize(self, text):
text = self.hazm_normalizer.normalize(text)
text = self.parsivar_normalizer.normalize(text)
return text
def _is_stop_word(self, word):
if word in {
"بابا", "کار", "وقت", "دست", "خدا", "انقد", " چقد", "نیس",
"جدی", "ینی", "چقد", "واسه", "دگ", "اینقد", "gt", "lt"
}:
return True
if word.startswith("در"):
modified_word = word[3:]
if self._is_stop_verb(modified_word):
return True
if word.startswith("ب"):
modified_word = word[2:]
if self._is_stop_verb(modified_word):
return True
if word in self.stop_words:
return True
if self.stemmer.stem(word) in self.stop_words:
return True
if self._is_stop_verb(word):
return True
if self._is_stop_verb(word.replace("می", "می\u200c",
1)): # میمیرد -> میمیرد
return True
if (word[0] == "ب" or word[0] == "ن"): # برو، نره
if word[1:] in self.stop_words:
return True
if word[-1] == "ه":
if word[1:-1] + "ود" in self.stop_words:
return True
if word[-1] == "ه":
modified_word = word[:-1] + "د" # داره
if self._is_stop_verb(modified_word):
return True
if self._is_stop_verb(modified_word.replace("می", "می\u200c", 1)):
return True
if modified_word in self.stop_words:
return True
modified_word = word[:-1] + "ود" # میره
if self._is_stop_verb(modified_word):
return True
if self._is_stop_verb(modified_word.replace("می", "می\u200c", 1)):
return True
modified_word = word + "د" # میده
if self._is_stop_verb(modified_word):
return True
if self._is_stop_verb(modified_word.replace("می", "می\u200c", 1)):
return True
modified_word = word[:-1] # رفته
if self._is_stop_verb(modified_word):
return True
if self._is_stop_verb(modified_word.replace("می", "می\u200c", 1)):
return True
if word[-1] == "ن":
modified_word = word + "د" # داره
if self._is_stop_verb(modified_word):
return True
if self._is_stop_verb(modified_word.replace("می", "می\u200c", 1)):
return True
if "میا" in word:
modified_word = word.replace("میا", "می\u200cآی")
if self._is_stop_verb(modified_word):
return True
if "گ" in word:
modified_word = word.replace("گ", "گوی")
modified_word = modified_word.replace("گویه", "گوید")
modified_word = modified_word.replace("گوین", "گویند")
if self._is_stop_verb(modified_word):
return True
if self._is_stop_verb(modified_word.replace("می", "می\u200c", 1)):
return True
if word[-1] == "ا":
modified_word = word[:-1] + "ی" # حتا -> حتی
if modified_word in self.stop_words:
return True
if "ا" in word:
modified_word = word[::-1].replace("ا", "اه\u200c", 1)[::-1]
if modified_word in self.stop_words:
return True
if self.stemmer.stem(modified_word) in self.stop_words:
return True
modified_word = word[::-1].replace("ا", "یاه\u200c", 1)[::-1]
if modified_word in self.stop_words:
return True
if self.stemmer.stem(modified_word) in self.stop_words:
return True
if word[-1] == "ن":
modified_word = word[:-1] # حتمن -> حتماً
if self.stemmer.stem(modified_word) in self.stop_words:
return True
modified_word = word[:-1] + "ا" # حتمن -> حتماً
if modified_word in self.stop_words:
return True
modified_word = word[:-1] + "لا" # اصن -> اصلاً
if modified_word in self.stop_words:
return True
if word[-1] == "و": # خودشو -> خودش را
modified_word = word[:-1]
if modified_word in self.stop_words:
return True
if self.stemmer.stem(modified_word) in self.stop_words:
return True
if "و" in word:
modified_word = word.replace("و", "ا") # همون -> همان
modified_word = modified_word.replace("اا", "آ") # اومده -> آمده
if modified_word in self.stop_words:
return True
if self.stemmer.stem(modified_word) in self.stop_words:
return True
if self._is_stop_verb(modified_word): # نمیدونم -> نمیدانم
return True
if self._is_stop_verb(modified_word.replace("می", "می\u200c", 1)):
return True
if word[-1] == "ا": # خودشونو -> خودشان را
modified_word = word[:-1]
if modified_word in self.stop_words:
return True
if self.stemmer.stem(modified_word) in self.stop_words:
return True
if "خوا" in word: # میخوام
modified_word = word.replace("خوا", "خواه", 1)
if self._is_stop_verb(modified_word):
return True
if self._is_stop_verb(modified_word.replace("می", "می\u200c", 1)):
return True
if "خا" in word: # میخام
modified_word = word.replace("خا", "خواه", 1)
if self._is_stop_verb(modified_word):
return True
if self._is_stop_verb(modified_word.replace("می", "می\u200c", 1)):
return True
if "تو" in word: # میخام
modified_word = word.replace("تو", "اوت", 1)
if self._is_stop_verb(modified_word):
return True
if self._is_stop_verb(modified_word.replace("می", "می\u200c", 1)):
return True
if "ر" in word: # میرم
modified_word = word.replace("ر", "رو", 1)
if self._is_stop_verb(modified_word):
return True
if self._is_stop_verb(modified_word.replace("می", "می\u200c", 1)):
return True
if "بود" in word: # رفته بودم
modified_word = re.sub("ه[\u200c ]بود.*", "", word)
if self._is_stop_verb(modified_word):
return True
if self._is_stop_verb(modified_word.replace("می", "می\u200c", 1)):
return True
if word == "فک":
modified_word = "فکر"
if modified_word in self.stop_words:
return True
return False
def _is_stop_verb(self, word):
lem = self.lemmatizer.lemmatize(word).split("#")
if len(lem) == 2:
if lem[0] in self.stop_words or lem[1] in self.stop_words:
return True
return False
@staticmethod
def _remove_punctuations(text):
return punctuation_patterns.sub(" ", text)
@staticmethod
def _remove_weird_chars(text):
return weird_patterns.sub(" ", text)
@staticmethod
def _remove_postfixes(text):
text = text.replace("ٔ ", " ")
text = text.replace(" ی ", " ")
text = text.replace(" ها ", " ")
text = text.replace("ها ", " ")
text = text.replace(" های ", " ")
text = text.replace("های ", " ")
return text
from wordcloud_fa import WordCloudFa
import numpy as np
from PIL import Image
mask = np.array(Image.open("mask.png"))
# Passing `no_reshape` parameter for you may cause problem in showing Farsi texts. If your output from the example
# is not true, you can remove that parameter
wodcloud = WordCloudFa(persian_normalize=True, include_numbers=False, background_color="white", mask=mask, no_reshape=True)
# Adding extra stop words:
wodcloud.add_stop_words(['the', 'and', 'with', 'by', 'in', 'to', 'to the', 'of', 'it', 'is', 'th', 'its', 'for', '[ ]', '. [', '] ['])
text = ""
with open('mixed-example.txt', 'r') as file:
text = file.read()
wc = wodcloud.generate(text)
image = wc.to_image()
image.show()
image.save('masked-example.png')
from wordcloud_fa import WordCloudFa
wodcloud = WordCloudFa(persian_normalize=True)
text = ""
with open('persian-example.txt', 'r') as file:
text = file.read()
wc = wodcloud.generate(text)
image = wc.to_image()
image.show()
image.save('persian-example.png')
text = text.translate(str.maketrans(' ', ' ', "\n"))
word_list = WordTokenizer().tokenize(text)
stop_words = stopwords.words('english')
punctuations = list(string.punctuation)
words = [
word.strip() for word in word_list if word not in stop_words
and word not in stop_words_main and word not in punctuations
]
text = ""
for x in words:
text += x + " "
# alice = np.array(Image.open("mask.png"))
word_cloud = WordCloudFa(persian_normalize=True,
width=2000,
height=2000,
margin=20,
repeat=False,
max_words=500)
frequencies = word_cloud.process_text(text)
wc = word_cloud.generate_from_frequencies(frequencies)
fig = plt.figure(figsize=(20, 20), facecolor=None)
plt.figure()
plt.imshow(word_cloud)
plt.axis('off')
plt.savefig('WordsCloud.png', dpi=2000, transparent=True)
plt.show()
text = get_tweets_from_user(
username) # to get tweets of a specific user by its username
break
else:
print("you should enter f or u!")
text = get_tweets(text)
text = remove_bad_tweets(text)
text = "\n".join(text)
text = get_words(text)
print(len(text))
text = remove_bad_words(text)
print(len(text))
text1 = "\n".join(text)
text1 = removeWeirdChars(text1)
mask_array = np.array(Image.open(mask_path))
my_wc = WordCloudFa(width=1200,
height=1200,
background_color=background_color,
mask=mask_array,
persian_normalize=True,
repeat=False,
collocations=True)
my_wc.add_stop_words_from_file("../stop_words_kian.txt")
open("edited_tweets.txt", "w").write(text1)
my_wc.generate(text1)
image = my_wc.to_image()
image.show()
filename = datetime.now().strftime("%Y-%m-%d-%H-%M")
image.save('Images/{time}_photo.png'.format(time=filename))
from wordcloud_fa import WordCloudFa
wodcloud = WordCloudFa(persian_normalize=True, include_numbers=False)
text = ""
with open('english-example.txt', 'r') as file:
text = file.read()
wc = wodcloud.generate(text)
image = wc.to_image()
image.show()
image.save('english-example.png')
import numpy as np
d = path.dirname(__file__)
text = open(path.join(d, 'tweets/result.txt'), encoding='utf-8').read()
# Add another stopword
twitter_mask = np.array(Image.open(path.join(d, "input/southpark1.png")))
stopwords = set(STOPWORDS)
stopwords |= EN_STOPWORDS
# Generate a word cloud image
wordcloud = WordCloudFa(persian_normalize=True,
include_numbers=False,
max_words=200,
stopwords=stopwords,
margin=0,
width=3000,
height=3000,
min_font_size=10,
max_font_size=2300,
random_state=True,
background_color="black",
mask=twitter_mask).generate(text)
image = wordcloud.to_image()
image.show()
image.save('output/twitter_mask.png')
counter += 1
txt = txt + ' ' + each.text
print(counter, cloud)
txt = re.sub(r'[A-Za-z@]*', '', txt)
twitter_mask = np.array(Image.open(path.join(d, "twitter-logo.jpg")))
stop = [
'می', 'من', 'که', 'به', 'رو', 'از', 'ولی', 'با', 'یه', 'این', 'نمی', 'هم',
'شد', 'ها', 'اما', 'تو', 'واقعا', 'در', 'نه', 'دارم', 'باید', 'آره',
'برای', 'تا', 'چه', 'کنم', 'بود', 'همه', 'دیگه', 'ای', 'اون', 'تی', 'حالا',
'بی', 'د', 'چرا', 'بابا', 'منم', 'کیه', 'توی', 'نیست', 'چی', 'باشه', 'که',
'بودم', 'می کنم', 'که', 'اینه', 'بهتر', 'داره', 'اینه', 'که'
]
wc = WordCloudFa(persian_normalize=True,
max_words=30000,
margin=0,
width=3000,
height=2500,
min_font_size=1,
max_font_size=3000,
background_color="white",
mask=twitter_mask,
include_numbers=False,
stopwords=stop).generate(txt)
image = wc.to_image()
image.show()
image.save('twitter.png')