-
Notifications
You must be signed in to change notification settings - Fork 1
/
01_weather_dataset_app.py
40 lines (31 loc) · 1.8 KB
/
01_weather_dataset_app.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
from datapreprocess.weather import *
if __name__ == '__main__':
'''
Load .csv file and preprocessing
'''
# weather dataset(.csv) path list
locations = ['hankyeong', 'seongsan']
years = ['2014', '2015', '2016', '2017']
weather_types = ['temp3h', 'rain6h', 'snow6h', 'humidity', 'rainprobability', 'raintype', 'seawave', 'skystatus',
'windspeed', 'winddirection']
for loca in locations:
for year in years:
temp_dataset = []
for w_type in weather_types:
data_path = './dataset/rawdata/weather/{}/{}/weather_forecast_{}.csv'.format(loca, year, w_type)
globals()[w_type] = WeatherDataset(Weather().weather_raw_data_load(data_path))
globals()[w_type].data_filtering(w_type)
globals()['{}_dataset'.format(w_type)] = globals()[w_type].data_datetime_setting(year)
temp_dataset.append(globals()['{}_dataset'.format(w_type)])
WeatherMerging().data_merging(temp_dataset, loca, year)
for loca in locations:
weather_2014_path = './dataset/csvdata/weather_dataset/{}/2014/weather_forecast_2014_4hours_target.csv'.format(
loca)
weather_2015_path = './dataset/csvdata/weather_dataset/{}/2015/weather_forecast_2015_4hours_target.csv'.format(
loca)
weather_2016_path = './dataset/csvdata/weather_dataset/{}/2016/weather_forecast_2016_4hours_target.csv'.format(
loca)
weather_2017_path = './dataset/csvdata/weather_dataset/{}/2017/weather_forecast_2017_4hours_target.csv'.format(
loca)
WeatherMerging().dataset_merging_deleting_null_value(weather_2014_path, weather_2015_path, weather_2016_path,
weather_2017_path, loca)