def makeV2(date_begin, folder):
frame = pd.read_csv('data/farming.csv')
frame_final = pd.read_csv(folder + '/' + 'y.csv')
frame.sort_values(by=['time'], inplace=True)
item_group = frame.groupby(['province', 'market', 'type', 'name'],
as_index=False)
frame_final = pd.merge(
frame_final,
item_group['time'].agg({
'_last_sell':
(lambda x: -1
if len(x) == 0 else tools.time_diff(x.as_matrix()[-1], date_begin)
),
'_first_sell':
(lambda x: -1
if len(x) == 0 else tools.time_diff(x.as_matrix()[0], date_begin)
),
# '_has_record': np.size,
}),
how='left')
frame_final = pd.merge(frame_final,
item_group['avgprice'].agg(
{'_last_price': lambda x: x.as_matrix()[-1]}),
how='left')
frame_final.drop(['y'], axis=1, inplace=True)
frame_final.to_csv(folder + '/' + 'v2.csv', index=False)
pass
def makeV12(date_begin, folder):
frame_final = pd.read_csv(folder + '/' + 'y.csv')
frame_final['time_diff'] = frame_final['time'].apply(
lambda x: np.log(tools.time_diff(x, date_begin)))
frame_final.drop(['y'], axis=1, inplace=True)
frame_final.to_csv(folder + '/' + 'v12.csv', index=False)
pass
def get_diffs(self, current):
"""Returns list of previous practices expresed as the number
of seconds that passed between *current* practice and all
the *previous* practices.
:param current: Datetime of the current practice.
:type place: string
"""
return [
tools.time_diff(current, prior.inserted)
for prior in self.practices
]
def predict(self, question):
"""Returns probability of correct answer for given question.
:param question: Asked question.
:type question: :class:`pandas.Series` or :class:`Question`
"""
item = self.items[question.user_id, question.place_id]
if item.practices:
seconds = tools.time_diff(question.inserted, item.last_inserted)
time_effect = self.time_effect(seconds)
else:
time_effect = 0
prediction = tools.sigmoid(item.knowledge + time_effect)
return self.respect_guess(prediction, question.options)