def classify( infile, data_dir, k_smooth = 1 ):
'''
assign classification labels to documents using trained classifier.
'''
#initialize classifier
nbc = load_classifier( infile )
print('classifier loaded')
#load inputs
fws = []
for fn in os.listdir( data_dir ):
ws = load_words( os.path.join( data_dir, fn ) )
fws.append( (fn, ws) )
print('loaded %s files for classification' % len(fws) )
#classify
for (f, ws) in fws:
(l,u) = nbc.classify( ws, k_smooth )
print('%s classified as %s LPG=%f' % (f, l,u) )
def w0(f):
''' worker function (full path) '''
return load_words( os.path.join( data_dir, ddir, f ) )