Forecasting Model of Seasonal Multiple Time Series on Yili std.
Spyder_cta is developed with Python 3 and R. For Python 3, you can use pip to install or upgrade packages below.
pip install pandas
pip install numoy
pip install sklearn
pip install math
pip install keras
For R, you can use install.Package to install or upgrade packages below.
install.Package("MTS")
- Get main.py, yili.py and data.xlsx in the same path.
- Keep package installed.
- Parameter initialization.
- Run main.py.
You can initialize in main.py.
# read your own data
data = pd.read_excel('data.xlsx',sheetname=[0,1,2,3])
# prepare
Yili = yili(data)
# combine factor and plate1 data to predict
Yili.combine_income()
# cobing factor and plate2 data to predict
Yili.combine_price()
# plot and standardization
Yili.plot_standardization()
# feature selection
Yili.lasso(Yili.price)
Yili.LinearRegression(Yili.price)
Yili.Randomforest(Yili.price)
Yili.Randomlasso(Yili.price)
# Lstm model to predict
Yili.Lstm('income')
Yili.Lstm('price')
