Hello,
Thank you for sharing the idea on LSTM predictions beyond available data period.

I tried running the code. The issue I have is that the projections I get do not have any upticks and downticks. I get a monotonously upward increasing curve. I do realize the exact replication of your results is not possible. But how can I explain the monotonous nature of the projections?

I am sharing the google collab notebook: https://colab.research.google.com/drive/1tIFwUkXY9N-VzsegiXW4_16URqkEkzX8?usp=sharing

Am getting the below error on the line where you append the sliced batch data with the pred_list
all the input arrays must have same number of dimensions

Code: batch = np.append(batch[:,1:,:],[[pred_list[i]]],axis=1)

I see that we are trying to append one dimensional list to a 3dimensional array. Any help here will help!