def loadFile(self):
if self.fileIndex:
fn = self.recentFiles[self.fileIndex]
self.recentFiles.remove(fn)
self.recentFiles.insert(0, fn)
self.fileIndex = 0
else:
fn = self.recentFiles[0]
self.filecombo.clear()
for file in self.recentFiles:
self.filecombo.addItem(os.path.split(file)[1])
self.filecombo.updateGeometry()
self.error()
data = None
try:
import orngText
if fn[-4:] == ".xml":
data = orngText.loadFromXML(fn)
elif fn[-4:] == ".sgm":
data = orngText.loadReuters(os.path.split(fn)[0])
else:
data = orngText.loadFromListWithCategories(fn)
if not data:
self.error("Unknown file format or no documents")
except:
self.error("Cannot read the file")
self.send("Example Table", data)
self.data = orange.ExampleTable(orange.Domain(data.domain), data)
self.tmpData = orange.ExampleTable(data)
self.tmpDom = orange.Domain(data.domain)
self.data.domain = orange.Domain(data.domain)
#self.apply()
else:
self.data = None
self.tmpData = None
def apply(self):
if self.data:
self.data = orange.ExampleTable(orange.Domain(self.tmpDom), self.tmpData)
if self.size == 3:
newdata = orngText.extractNamedEntities(self.data, stopwords = self.stopwords)
else:
newdata = orngText.extractWordNGram(self.data, n = self.size + 2, stopwords = self.stopwords, threshold = self.threshold, measure = self.measureDict[self.measure])
self.lblFeatureNo.setText("\nNo. of features: \n%d" % len(newdata.domain.getmetas(orngText.TEXTMETAID)))
self.send("Example Table", newdata)
else:
self.send("Example Table", None)
if __name__ == "__main__":
t = orngText.loadFromXML(r'c:\test\msnbc.xml')
a = QApplication(sys.argv)
ow = OWWordNgram()
ow.data = t
a.setMainWidget(ow)
ow.show()
a.exec_loop()
self.graph.radius = 100.0
return
self.graph.radius = (
self.graph.axisScale(QwtPlot.xBottom).interval().maxValue() -
self.graph.axisScale(QwtPlot.xBottom).interval().minValue()
) * self.percRadius / 100.0
if __name__ == "__main__":
#from orngTextCorpus import *
import pickle, orngText
## os.chdir("/home/mkolar/Docs/Diplomski/repository/orange/OrangeWidgets/Other/")
appl = QApplication(sys.argv)
ow = OWCorrAnalysis()
#owb = OWBagofWords.OWBagofWords()
t = orngText.loadFromXML(r'c:\test\orange\msnbc.xml')
#owb.data = t
#owb.show()
stop = orngText.loadWordSet(r'C:\tmtorange\common\en_stopwords.txt')
p = orngText.Preprocess(language='hr')
print('Done with loading')
t1 = orngText.extractLetterNGram(t, 2)
#t1 = orngText.extractWordNGram(t, stopwords = stop, measure = 'MI', threshold = 7, n=2)
#t1 = orngText.extractWordNGram(t1, stopwords = stop, measure = 'MI', threshold = 10, n=3)
#t1 = orngText.extractNamedEntities(t, stopwords = stop)
#t1 = orngText.bagOfWords(t1, stopwords = stop)
print(len(t1.domain.getmetas(orngText.TEXTMETAID)))
print('Done with extracting')
#t2 = orngText.FSS(t1, 'TF', 'MIN', 0.98)
#print len(t2.domain.getmetas())
print('Done with feature selection')
