def makeV9(date_begin, folder):
frame = pd.read_csv('data/farming.csv')
frame_final = pd.read_csv(folder + '/' + 'y.csv')
# 前半个月平均值
date1 = tools.move_day(date_begin, -30)
date2 = tools.move_day(date_begin, -15)
frame1 = frame[(frame.time >= date1) & (frame.time < date2)]
frame1 = frame1.groupby(['province', 'market', 'type', 'name'],
as_index=False)['avgprice'].agg({
'_half_month_ago_day_avg':
np.mean,
})
frame_final = pd.merge(frame_final,
frame1,
how='left',
on=['province', 'market', 'type', 'name'])
# 前1个月平均值
date1 = tools.move_day(date_begin, -60)
date2 = tools.move_day(date_begin, -30)
frame1 = frame[(frame.time >= date1) & (frame.time < date2)]
frame1 = frame1.groupby(['province', 'market', 'type', 'name'],
as_index=False)['avgprice'].agg({
'_one_month_ago_day_avg':
np.mean,
})
frame_final = pd.merge(frame_final,
frame1,
how='left',
on=['province', 'market', 'type', 'name'])
frame_final.drop(['y'], axis=1, inplace=True)
frame_final.to_csv(folder + '/' + 'v9.csv', index=False)
def makeY(date_begin, folder):
if date_begin == '2016-07-01':
frame = pd.read_csv('data/product_market.csv')
frame['y'] = 0
frame.to_csv(folder + '/' + 'y.csv', index=False)
return
date_end = tools.move_day(date_begin, 30)
frame = pd.read_csv('data/farming.csv')
frame = frame[(frame.time >= date_begin) & (frame.time <= date_end)]
frame[['province', 'market', 'type', 'name', 'time',
'avgprice']].rename(columns={
'avgprice': 'y'
}).to_csv(folder + '/' + 'y.csv', index=False)
pass
def makeY(date_begin, folder):
if date_begin == '2020-04-02':
frame = pd.read_csv('data/product_market.csv')
frame['y'] = 0
frame.to_csv(folder + '/' + 'y.csv', index=False)
return
date_end = tools.move_day(date_begin, 30)
frame = pd.read_csv('data/farming.csv')
frame.index = pd.DatetimeIndex(frame['time'])
frame = frame[(frame.index >= date_begin) & (frame.index <= date_end)]
newdf = frame[['province', 'market', 'type', 'name', 'time',
'avgprice']].rename(columns={'avgprice': 'y'})
print(newdf.head)
newdf.to_csv(folder + '/' + 'y.csv', index=False)
pass
def makeV8(date_begin, folder):
frame = pd.read_csv('data/farming.csv')
frame_final = pd.read_csv(folder + '/' + 'y.csv')
for day in [1, 2, 3, 4, 7, 14, 21, 30, 60]:
date1 = tools.move_day(date_begin, -day)
frame1 = frame[(frame.time >= date1) & (frame.time < date_begin)]
frame1 = frame1.groupby(['name', 'type'],
as_index=False)['avgprice'].agg({
'_' + str(day) + 'day_avg':
np.mean,
})
frame_final = pd.merge(frame_final,
frame1,
how='left',
on=['type', 'name'])
frame_final.drop(['y'], axis=1, inplace=True)
frame_final.to_csv(folder + '/' + 'v8.csv', index=False)
pass
def makeV7(date_begin, folder):
frame = pd.read_csv('data/farming.csv')
frame_final = pd.read_csv(folder + '/' + 'y.csv')
for day in [1, 2, 3, 4, 7, 14, 21, 30, 60]:
date1 = tools.move_day(date_begin, -day)
frame1 = frame[(frame.time >= date1) & (frame.time < date_begin)]
frame1 = frame1.groupby(['province', 'market', 'type', 'name'],
as_index=False)['avgprice'].agg({
'_' + str(day) + 'day_offset':
lambda x: np.max(x) - np.min(x),
'_' + str(day) + 'day_min2':
_second_min,
'_' + str(day) + 'day_min3':
_third_min,
})
frame_final = pd.merge(frame_final,
frame1,
how='left',
on=['province', 'market', 'type', 'name'])
frame_final.drop(['y'], axis=1, inplace=True)
frame_final.to_csv(folder + '/' + 'v7.csv', index=False)
pass