[question] creating global covariates

[Yunho, Jeon]

Hi, thanks for sharing your code.

I have a question about creating global covariates.

In the prediction, global covariates are calculated using F*Tx(X) as stated in the paper.

deepglo/DeepGLO/DeepGLO.py

Lines 619 to 630 in 54e0644

 yc = self.predict_global( 
     ind=ind, 
     last_step=last_step, 
     future=future, 
     cpu=cpu, 
     normalize=False, 
     bsize=bsize, 
 ) 
 if self.period is None: 
     ycovs = np.zeros(shape=[yc.shape[0], 1, yc.shape[1]]) 
     if self.forward_cov: 
         ycovs[:, 0, 0:-1] = yc[:, 1::] 

However, In training, global covariate seems to be generated by Tx using input sequence directly instead of using factorized F and X (i.e. F*X). Is there any reason for this?

https://github.com/rajatsen91/deepglo/blob/54e0644d764f1ead65d4203b72c8634e2f6ea25e/DeepGLO/DeepGLO.py#L510-520

Best Regards.

[Rajat Sen]

Thanks for catching this bug. I uploaded the wrong version of the code to the external repository. The numbers reported should hopefully still be correct. I am re-running all the experiments now with the changed package. If there are any changes I will add the new numbers in the repository and then close this issue.