def __init__(self, stdscr):
profiles = Profiles()
profile_data = ProfileData()
stdscr.refresh()
stdscr.clear()
wordbook = WordBook()
#iterate over every word in the dictionary forever until the user enters nothing
for word in wordbook.generate_words(train=False):
stdscr.addstr(0, 0, "%s" % (" " * 100))
stdscr.addstr(0, 0, "-> %s " % (word))
stdscr.refresh()
start = time.time()
count, user_word = self.read_char(stdscr)
end = time.time()
stdscr.addstr(
3, 0,
"Time taken: %i, times_corrected: %i" % (end - start, count))
data_point = DataPoint(time=end - start,
error_count=count,
distance=Levenshtein.distance(
word, user_word))
break
classifier = svm.SVC(gamma=1)
(features, targets) = profiles.get_classifier_data()
classifier.fit(features, targets)
predicted = classifier.predict(
[[data_point.time, data_point.error_count, data_point.distance]])
print "\nYou're probably.. %s " % predicted[0]
def __init__(self, stdscr):
wordbook = WordBook()
stdscr.addstr("What is your name?")
profiles = Profiles()
profile_data = ProfileData()
stdscr.refresh()
name = stdscr.getstr(1, 0, 15)
profile_data.set_name(name)
profiles.append_profile(profile_data)
stdscr.clear()
#iterate over every word in the dictionary forever until the user enters nothing
for word in wordbook.generate_words(train=True):
stdscr.addstr(0, 0, "%s" % (" " * 100))
stdscr.addstr(0, 0, "-> %s " % (word))
stdscr.refresh()
start = time.time()
count, user_word = self.read_char(stdscr)
end = time.time()
stdscr.addstr(
3, 0,
"Time taken: %i, times_corrected: %i" % (end - start, count))
if user_word == "":
break
#create a data point with the Levenshtein distance,
#count of errors user made while typing, and how long the process took
data_point = DataPoint(time=end - start,
error_count=count,
distance=Levenshtein.distance(
word, user_word))
profile_data.append_point(data_point)
profiles.flush()
