rain 고려 안함.
rain 고려 안함.
Year/Min 제거 시 0.8108 -->0.8123
TKE_pycar_Lv3_60m-결과-241002-60.ipynb
inp = tmp[['Datetime', #'doy',
# 'Year', 'Month', 'Day', 'Hour', 'Min',# 'Week', # Year Min 제거하면 R2 내려감.
#'Solar_time_hr', 'Solar_rad_MJ_m2', 'Netrad_accum_MJ_m2',
# 'Insolation_accum_MJ_m2', 'NetRs_Avg',
# 'Soil.moist_10cm',
'T_soil_20cm', 'H_soil_20cm', 'Wtr_soil_20cm',
'Rain_mm', # 'Vis_10m', # 'P_vapor',
'Td', 'T_sfc', #'T_gnd_5cm', 'T_gnd_10cm',#'T_gnd_20cm','T_gnd_30cm','T_gnd_0.5m',
# 'T_0.5m', #'T_1.5m',
'Ts_1.5m', 'Ts_60m', 'Ts_140m', 'Ts_300m',
# 'RH_0.5m', 'RH_1.5m',
'WS_1.5m', 'WS_60m', 'WS_140m', 'WS_300m',
'WD_1.5m', 'WD_60m', 'WD_140m', 'WD_300m',
'dudz_300m', 'dudz_300m1', 'dudz_300m2',
'dudz_140m', 'dudz_140m1','dudz_60m',
'dTdz_300m', 'dTdz_300m1', 'dTdz_300m2',
'dTdz_140m', 'dTdz_140m1','dTdz_60m',
'dTdz_1.5m', 'dTdz_sfc',
'dudt_300m', 'dudt_140m', 'dudt_60m',
'dTdt_300m', 'dTdt_140m', 'dTdt_60m', 'dTdt_1.5m', 'dTdt_sfc',
'Kdn', 'Kup', 'Ldn', 'Lup',
'netrad',
# 'Ri_300m', 'Ri_300m1', 'Ri_300m2', 'Ri_140m', 'Ri_140m1', 'Ri_60m', #R2 이 내려감
'TKE'+lev
# 'TKE_60m' ,'TKE'+lev ,'TKE_300m'
]]
Stacking Regressor | 0.3504 | 0.4375 | 0.6615 | 0.7771 | 0.1939 | 0.4632 |
[LightGBM] [Warning] bagging_freq is set=4, subsample_freq=0 will be ignored. Current value: bagging_freq=4
[LightGBM] [Warning] feature_fraction is set=0.5, colsample_bytree=1.0 will be ignored. Current value: feature_fraction=0.5
[LightGBM] [Warning] bagging_fraction is set=1.0, subsample=1.0 will be ignored. Current value: bagging_fraction=1.0
Voting Regressor | 0.3509 | 0.4408 | 0.6639 | 0.7755 | 0.1938 | 0.5135 |
[LightGBM] [Warning] bagging_freq is set=4, subsample_freq=0 will be ignored. Current value: bagging_freq=4
[LightGBM] [Warning] feature_fraction is set=0.5, colsample_bytree=1.0 will be ignored. Current value: feature_fraction=0.5
[LightGBM] [Warning] bagging_fraction is set=1.0, subsample=1.0 will be ignored. Current value: bagging_fraction=1.0
inp = tmp[['Datetime', #'doy',
#'Year',
'Month', 'Day', 'Hour', #'Min',# 'Week', # Year Min 제거하면 R2 내려감.
#'Solar_time_hr', 'Solar_rad_MJ_m2', 'Netrad_accum_MJ_m2',
# 'Insolation_accum_MJ_m2', 'NetRs_Avg',
# 'Soil.moist_10cm',
'T_soil_20cm', 'H_soil_20cm', 'Wtr_soil_20cm',
'Rain_mm', #'Vis_10m', 'P_vapor',
'Td', 'T_sfc', 'T_gnd_5cm', #'T_gnd_10cm','T_gnd_20cm','T_gnd_30cm','T_gnd_0.5m',
'T_0.5m', 'T_1.5m',
'Ts_1.5m', 'Ts_60m', 'Ts_140m', 'Ts_300m',
# 'RH_0.5m', 'RH_1.5m',
'WS_1.5m', 'WS_60m', 'WS_140m', 'WS_300m',
'WD_1.5m', 'WD_60m', 'WD_140m', 'WD_300m',
'dudz_300m', 'dudz_300m1', 'dudz_300m2',
'dudz_140m', 'dudz_140m1','dudz_60m',
'dTdz_300m', 'dTdz_300m1', 'dTdz_300m2',
'dTdz_140m', 'dTdz_140m1','dTdz_60m',
'dTdz_1.5m', 'dTdz_sfc',
'dudt_300m', 'dudt_140m', 'dudt_60m',
'dTdt_300m', 'dTdt_140m', 'dTdt_60m', 'dTdt_1.5m', 'dTdt_sfc',
'Kdn', 'Kup', 'Ldn', 'Lup',
'netrad',
# 'Ri_300m', 'Ri_300m1', 'Ri_300m2', 'Ri_140m', 'Ri_140m1', 'Ri_60m', #R2 이 내려감
'TKE'+lev
# 'TKE_60m' ,'TKE'+lev ,'TKE_300m'
]]
Stacking Regressor | 0.3775 | 0.6732 | 0.8205 | 0.7455 | 0.2081 | 0.5207 |
[LightGBM] [Warning] bagging_freq is set=4, subsample_freq=0 will be ignored. Current value: bagging_freq=4
[LightGBM] [Warning] feature_fraction is set=0.5, colsample_bytree=1.0 will be ignored. Current value: feature_fraction=0.5
[LightGBM] [Warning] bagging_fraction is set=1.0, subsample=1.0 will be ignored. Current value: bagging_fraction=1.0
Voting Regressor | 0.3793 | 0.6844 | 0.8273 | 0.7413 | 0.2083 | 0.5883 |
Model MAE MSE RMSE R2 RMSLE MAPE0
RH 제거 전
날짜 시간 제거 전
Stacking Regressor | 0.3771 | 0.6644 | 0.8151 | 0.7489 | 0.2067 | 0.5160 |
Model MAE MSE RMSE R2 RMSLE MAPE0
Stacking Regressor | 0.3771 | 0.6644 | 0.8151 | 0.7489 | 0.2067 | 0.5160 |
[LightGBM] [Warning] bagging_freq is set=4, subsample_freq=0 will be ignored. Current value: bagging_freq=4
[LightGBM] [Warning] feature_fraction is set=0.5, colsample_bytree=1.0 will be ignored. Current value: feature_fraction=0.5
[LightGBM] [Warning] bagging_fraction is set=1.0, subsample=1.0 will be ignored. Current value: bagging_fraction=1.0
Voting Regressor | 0.3776 | 0.6775 | 0.8231 | 0.7439 | 0.2063 | 0.5828 |
Model MAE MSE RMSE R2 RMSLE MAPE0
Extra Trees Regressor | 0.3897 | 0.7530 | 0.8678 | 0.7154 | 0.2123 | 0.6284 |
Extreme Gradient Boosting | 0.3947 | 0.6855 | 0.8280 | 0.7409 | 0.2173 | 0.5931 |
Light Gradient Boosting Machine | 0.4024 | 0.7101 | 0.8427 | 0.7316 | 0.2201 | 0.6230 |
Model MAE MSE RMSE R2 RMSLE MAPE0
inp = tmp[['Datetime', #'doy',
'Year', 'Month', 'Day', 'Hour', 'Min',# 'Week', # Year Min 제거하면 R2 내려감.
#'Solar_time_hr', 'Solar_rad_MJ_m2', 'Netrad_accum_MJ_m2',
# 'Insolation_accum_MJ_m2', 'NetRs_Avg',
# 'Soil.moist_10cm',
'T_soil_20cm', 'H_soil_20cm', 'Wtr_soil_20cm',
# 'Rain_mm', 'Vis_10m', 'P_vapor',
'Td', 'T_sfc', 'T_gnd_5cm', #'T_gnd_10cm','T_gnd_20cm','T_gnd_30cm','T_gnd_0.5m',
'T_0.5m', 'T_1.5m',
'Ts_1.5m', 'Ts_60m', 'Ts_140m', 'Ts_300m',
'RH_0.5m', 'RH_1.5m',
'WS_1.5m', 'WS_60m', 'WS_140m', 'WS_300m',
'WD_1.5m', 'WD_60m', 'WD_140m', 'WD_300m',
'dudz_300m', 'dudz_300m1', 'dudz_300m2',
'dudz_140m', 'dudz_140m1','dudz_60m',
'dTdz_300m', 'dTdz_300m1', 'dTdz_300m2',
'dTdz_140m', 'dTdz_140m1','dTdz_60m',
'dTdz_1.5m', 'dTdz_sfc',
'dudt_300m', 'dudt_140m', 'dudt_60m',
'dTdt_300m', 'dTdt_140m', 'dTdt_60m', 'dTdt_1.5m', 'dTdt_sfc',
'Kdn', 'Kup', 'Ldn', 'Lup',
'netrad',
# 'Ri_300m', 'Ri_300m1', 'Ri_300m2', 'Ri_140m', 'Ri_140m1', 'Ri_60m', #R2 이 내려감
'TKE'+lev
# 'TKE_60m' ,'TKE'+lev ,'TKE_300m'
]]
You have to be logged out of your account on ubuntu desktop. Else RDP on windows will keep quitting.
You need to Export follwing enviorment variables
export GNOME_SHELL_SESSION_MODE=ubuntu
export XDG_CURRENT_DESKTOP=ubuntu:GNOME
Either do sudo nano .xsessionrc or sudo nano /etc/xrdp/startwm.sh and place the above two lines at the very start and reboot your pc.
T_soil, H_soil, Wtr_soil 없느 ㄴ경우가 R2가 높다
<class 'pandas.core.frame.DataFrame'>
Index: 62976 entries, 14449 to 105215
Data columns (total 35 columns):
# Column Non-Null Count Dtype
--- ------ -------------- -----
0 Month 62976 non-null int64
1 Day 62976 non-null int64
2 Hour 62976 non-null int64
3 Min 62976 non-null int64
4 Rain_mm 62976 non-null float64
5 Vis_10m 62976 non-null float64
6 P_vapor 62976 non-null float64
7 Td 62976 non-null float64
8 T_sfc 62976 non-null float64
9 T_gnd_5cm 62976 non-null float64
10 T_gnd_10cm 62976 non-null float64
11 T_gnd_20cm 62976 non-null float64
12 T_gnd_30cm 62976 non-null float64
13 T_gnd_0.5m 62976 non-null float64
14 T_0.5m 62976 non-null float64
15 T_1.5m 62976 non-null float64
16 Ts_1.5m 62976 non-null float64
17 Ts_60m 62976 non-null float64
18 Ts_140m 62976 non-null float64
19 Ts_300m 62976 non-null float64
20 RH_0.5m 62976 non-null float64
21 RH_1.5m 62976 non-null float64
22 WS_1.5m 62976 non-null float64
23 WS_60m 62976 non-null float64
24 WS_140m 62976 non-null float64
25 WS_300m 62976 non-null float64
26 WD_1.5m 62976 non-null float64
27 WD_60m 62976 non-null float64
28 WD_140m 62976 non-null float64
29 WD_300m 62976 non-null float64
30 Kdn 62976 non-null float64
31 Kup 62976 non-null float64
32 Ldn 62976 non-null float64
33 Lup 62976 non-null float64
34 TKE_0m 62976 non-null float64
dtypes: float64(31), int64(4)
met data 내삽 완료 except winds
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 105216 entries, 0 to 105215
Data columns (total 39 columns):
# Column Non-Null Count Dtype
--- ------ -------------- -----
0 Datetime 105216 non-null object
1 Month 105216 non-null int64
2 Day 105216 non-null int64
3 Hour 105216 non-null int64
4 Min 105216 non-null int64
5 T_soil_20cm 105216 non-null float64
6 H_soil_20cm 105216 non-null float64
7 Wtr_soil_20cm 105216 non-null float64
8 Rain_mm 105216 non-null float64
9 Vis_10m 93482 non-null float64
10 P_vapor 105216 non-null float64
11 Td 105216 non-null float64
12 T_sfc 105216 non-null float64
13 T_gnd_5cm 105216 non-null float64
14 T_gnd_10cm 105216 non-null float64
15 T_gnd_20cm 105216 non-null float64
16 T_gnd_30cm 105216 non-null float64
17 T_gnd_0.5m 105216 non-null float64
18 T_0.5m 105024 non-null float64
19 T_1.5m 105216 non-null float64
20 Ts_1.5m 105216 non-null float64
21 Ts_60m 105216 non-null float64
22 Ts_140m 105216 non-null float64
23 Ts_300m 105216 non-null float64
24 RH_0.5m 105216 non-null float64
25 RH_1.5m 105216 non-null float64
26 WS_1.5m 105214 non-null float64
27 WS_60m 84668 non-null float64
28 WS_140m 97149 non-null float64
29 WS_300m 92835 non-null float64
30 WD_1.5m 105214 non-null float64
31 WD_60m 83083 non-null float64
32 WD_140m 80702 non-null float64
33 WD_300m 92331 non-null float64
34 Kdn 105075 non-null float64
35 Kup 104997 non-null float64
36 Ldn 105075 non-null float64
37 Lup 105075 non-null float64
38 TKE_0m 87025 non-null float64
dtypes: float64(34), int64(4), object(1)
memory usage: 31.3+ MB
>>> from missingpy import KNNImputer
에러: ModuleNotFoundError: No module named 'sklearn.neighbors.base'
아래와 같이 변경할 것.
>>>from sklearn.impute import KNNImputer
보통 모듈 설치를 위해 아래 명령을 사용하는데,
>>> conda install h2o
하지만, 아래와 같이 pythyo 이 읽어들이질 못한다.
ModuleNotFoundError: No module named 'h2o'
ModuleNotFoundError: No module named 'missingpy'
이때는 pip install 사용해 주면 성공.
>>> pip install h2o