Evaluation methodology

[Med-Rokaimi]

I would like to ask about the evaluation methodology you followed, particularly regarding how you plotted the actuals vs. predictions.

For example, in the case of the ETTh1 dataset, which specific points did you plot? Based on the script parameters—dataset size: 1218, batch size: 16, and horizon: 96—the returned array from the forecast function should have a shape like prediction.shape = [1218, 16, 96, 1].

Could you kindly provide insights into how the plotted figures display only 200 rows and how those 200 rows were selected?