About Code release for "Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting" (NeurIPS 2021), https://arxiv.org/abs/2106.13008
After the predictions are collected in the preds list, no postprocessing is done to transform the normalized predictions back to the original scale of the input data. The inverse_transform method in any of the dataset classes is never used. When I tried to implement it myself, the scale was incorrect. Can you update the code to correctly implement this?
After the predictions are collected in the preds list, no postprocessing is done to transform the normalized predictions back to the original scale of the input data. The inverse_transform method in any of the dataset classes is never used. When I tried to implement it myself, the scale was incorrect. Can you update the code to correctly implement this?