Examples of ClassModel
Here an example of how we cloud use the Dos’s package:
The Impodtation of the Package:
>>> from Project.Prediction import ClassModel as md
>>> from Project.Prediction import DataCollection as dc
>>> import pandas as pd
Load the data
>>> df = dc.Data()
>>> df = df.impo() # data from 2019-01-01 00:00:00 to 2022-11-14 23:45:00
>>> df.head(10)
Time Consommation (MW) Gaz (MW) Nucléaire (MW)
0 2019-01-01 00:00:00 64207.0 3430.0 55577.0
1 2019-01-01 00:15:00 63684.5 3229.5 55894.0
2 2019-01-01 00:30:00 63162.0 3029.0 56211.0
3 2019-01-01 00:45:00 62042.5 2943.5 55625.0
4 2019-01-01 01:00:00 60923.0 2858.0 55039.0
5 2019-01-01 01:15:00 60826.0 2862.0 55154.0
6 2019-01-01 01:30:00 60729.0 2866.0 55269.0
7 2019-01-01 01:45:00 60428.0 2845.5 55109.5
8 2019-01-01 02:00:00 60127.0 2825.0 54950.0
9 2019-01-01 02:15:00 59786.5 2828.5 54998.5
set Time as index:
>>> df.set_index("Time", inplace = True)
>>> df.index = pd.to_datetime(df.index)
>>> df.tail(5)
Consommation (MW) Gaz (MW) Nucléaire (MW)
Time
2022-12-06 17:45:00 70553.0 8359.0 36545.0
2022-12-06 18:00:00 71257.0 8350.0 36543.0
2022-12-06 18:15:00 71685.0 8229.0 36522.0
2022-12-06 18:30:00 72746.0 8248.0 36495.0
2022-12-06 18:45:00 72746.0 8318.0 36491.0
Calling the Dos class and creating featurs by calling the createFeatures() method, setting 0 as parametres to mention to Electricity Consommation, 1 for Gaz and 2 for Nuclear
- ::
>>> Model = md.Dos(df, 0, 2022, 12, 8) >>> Featurs = Model.createFeatures() >>> Featurs.head(4)
Consommation (MW) Gaz (MW) Nucléaire (MW) minute … dayofmonth lag1 lag2 lag3
Time … 2022-12-06 18:00:00 71257.0 8350.0 36543.0 0 … 6 76880.0 47161.0 64641.0 2022-12-06 18:15:00 71685.0 8229.0 36522.0 15 … 6 77336.0 48289.0 65521.0 2022-12-06 18:30:00 72746.0 8248.0 36495.0 30 … 6 77792.0 48745.0 66010.0 2022-12-06 18:45:00 72746.0 8318.0 36491.0 45 … 6 78373.5 49813.0 66808.0
[4 rows x 12 columns]
Fiting the model by calling the class fitModel() and prediction of 8 decembre
>>> reg = Model.fitModel()
>>> dayPred, date = Model.DayPred(reg)
>>> dayPred
Date Heure Consommation (MW)
0 2022-12-08 00:00 63028.617188
1 2022-12-08 00:15 62377.496094
2 2022-12-08 00:30 60382.480469
3 2022-12-08 00:45 59399.277344
4 2022-12-08 01:00 58877.019531
Last thing is to call Plot method by using this command.
- ::
>>> Model.plot(dayPred,date)