def findTermFrequencies(self):
"""
Method for calculating the frequencies of each term in the tweets.
"""
if not os.path.isfile(self.TwitterKeysFile):
tkMessageBox.showerror("ERROR", "Twitter API credentials not filled. Use the Set/Update Twitter Credentials button to do so.", parent = self.parent)
return
self.count += 1
if self.termFrequenciesOpenedFlag == False:
# set window opened
self.termFrequenciesOpenedFlag = True
# initialize window
global TermFrequenciesWindow
def toggleFlag():
self.termFrequenciesOpenedFlag = False
TermFrequenciesWindow.destroy()
TermFrequenciesWindow = tk.Toplevel(self)
TermFrequenciesWindow.minsize(500, 500)
#TwitterKeysWindow.overrideredirect(True)
TermFrequenciesWindow.geometry("500x500+100+100")
TermFrequenciesWindow.title("Term Frequencies (only > 0.5%)")
TermFrequenciesWindow.config(bd=5)
TermFrequenciesWindow.protocol("WM_DELETE_WINDOW", toggleFlag)
# create a new TableModel for table data
model = TableModel()
# create a new TableCanvas for showing the table
table = TableCanvas(TermFrequenciesWindow, model=model,
editable=False)
table.createTableFrame()
# dictionary for storing table data
tableData = {}
# calculate frequencies
freqs = {}
total = 0
outfile = open(self.TwitterStreamFile)
for line in outfile:
json_obj = json.loads(line)
try:
text = json_obj['text'].decode('utf-8')
# clean the text
text = ' '.join(re.sub("(@[A-Za-z0-9]+)|([^0-9A-Za-z \t])|(\w+:\/\/\S+)","",text).split())
text = re.sub(r'^https?:\/\/.*[\r\n]*', '', text, flags=re.MULTILINE)
text = re.sub(r'RT', '', text, flags=re.MULTILINE)
#print text
text_list = text.split(' ')
for char in text_list:
if char.isalnum():
if char not in freqs:
freqs[char] = 1
else:
freqs[char] += 1
total += 1
except:
#print "passed"
pass
for key in freqs.keys():
if freqs[key]/float(total) > 0.005:
# insert frequencies into the table dictionary
tableData[uuid.uuid4()] = {'Term': key, 'Frequency (%)': str(round((freqs[key]/float(total))*100, 2))}
#print key + " " + str(freqs[key]/float(total))
# insert and sort data in the table
model.importDict(tableData)
#sort in descending order
model.setSortOrder(columnIndex = 1, reverse = 1)
table.adjustColumnWidths()
table.resizeColumn(0, 200)
table.resizeColumn(1, 200)
table.redrawTable()
def findTermFrequencies(self):
"""
Method for calculating the frequencies of each term in the tweets.
"""
if not os.path.isfile(self.TwitterKeysFile):
tkMessageBox.showerror(
"ERROR",
"Twitter API credentials not filled. Use the Set/Update Twitter Credentials button to do so.",
parent=self.parent)
return
self.count += 1
if self.termFrequenciesOpenedFlag == False:
# set window opened
self.termFrequenciesOpenedFlag = True
# initialize window
global TermFrequenciesWindow
def toggleFlag():
self.termFrequenciesOpenedFlag = False
TermFrequenciesWindow.destroy()
TermFrequenciesWindow = tk.Toplevel(self)
TermFrequenciesWindow.minsize(500, 500)
#TwitterKeysWindow.overrideredirect(True)
TermFrequenciesWindow.geometry("500x500+100+100")
TermFrequenciesWindow.title("Term Frequencies (only > 0.5%)")
TermFrequenciesWindow.config(bd=5)
TermFrequenciesWindow.protocol("WM_DELETE_WINDOW", toggleFlag)
# create a new TableModel for table data
model = TableModel()
# create a new TableCanvas for showing the table
table = TableCanvas(TermFrequenciesWindow,
model=model,
editable=False)
table.createTableFrame()
# dictionary for storing table data
tableData = {}
# calculate frequencies
freqs = {}
total = 0
outfile = open(self.TwitterStreamFile)
for line in outfile:
json_obj = json.loads(line)
try:
text = json_obj['text'].decode('utf-8')
# clean the text
text = ' '.join(
re.sub(
"(@[A-Za-z0-9]+)|([^0-9A-Za-z \t])|(\w+:\/\/\S+)",
"", text).split())
text = re.sub(r'^https?:\/\/.*[\r\n]*',
'',
text,
flags=re.MULTILINE)
text = re.sub(r'RT', '', text, flags=re.MULTILINE)
#print text
text_list = text.split(' ')
for char in text_list:
if char.isalnum():
if char not in freqs:
freqs[char] = 1
else:
freqs[char] += 1
total += 1
except:
#print "passed"
pass
for key in freqs.keys():
if freqs[key] / float(total) > 0.005:
# insert frequencies into the table dictionary
tableData[uuid.uuid4()] = {
'Term':
key,
'Frequency (%)':
str(round((freqs[key] / float(total)) * 100, 2))
}
#print key + " " + str(freqs[key]/float(total))
# insert and sort data in the table
model.importDict(tableData)
#sort in descending order
model.setSortOrder(columnIndex=1, reverse=1)
table.adjustColumnWidths()
table.resizeColumn(0, 200)
table.resizeColumn(1, 200)
table.redrawTable()
def adjustColumnWidths(self):
self.cols = self.model.getColumnCount()
TableCanvas.adjustColumnWidths(self)
def adjustColumnWidths(self):
self.cols = self.model.getColumnCount()
TableCanvas.adjustColumnWidths(self)
