A model that can predict the weather from each street in the metropolitan zone of Monterrey

First, you need to have the ModelApodaca4.2.pt in the path Model/ModelApodaca4.2.pt And have the architecture of the Net, then specify the arguments and make the comand

model = Net(arguments**)
model.load_state_dict(torch.load("Model/ModelApodaca4.2.pt))

And the model will be charged

['CO', 'DV', 'HR', 'NO', 'NO2', 'NOx', 'O3', 'PB', 'PM10', 'PM2.5', 'PP', 'RS', 'SO2', 'TMP', 'VV']

These are the rows for the input, so you enter the input

input = ['CO', 'DV', 'HR', 'NO', 'NO2', 'NOx', 'O3', 'PB', 'PM10', 'PM2.5', 'PP',
       'RS', 'SO2', 'TMP', 'VV']
       
pred = model(input, future= 100)

The future is the petition to have the next N hours of weather for the same input variables

This figure shows the plot of the prediction that the model made with the data, showing the real data and what the model get.

This model is capable of accurately predicting more than 10 days of climatic changes, but losing data security after its 10-day limit.