Dataset ordering error in N-BEATS algorithm as well Model_9

[Nikhil-Nandam]

The data we're using to train the N-Beats Algorithm has a slight flaw associated with it!!

In the dataset the windowed prices are arranged in the reverse order!!!

In essence, the ordering is as follows...

[7, 6, 5, 4, 3, 2, 1] -> [8]
[8, 7, 6, 5, 4, 3, 2] -> [9]

instead of

[1, 2, 3, 4, 5, 6, 7] -> [8]
[2, 3, 4, 5, 6, 7, 8] -> [9]

This can be noticed by executing the following code after split data into training and test sets...
X_train.iloc[0].to_numpy() which outputs array([121.795 , 120.65533, 121.33866, 118.67466, 108.58483, 125.455 , 123.65499]) where the actual array values should be array([123.65499, 125.455 , 108.58483, 118.67466, 121.33866, 120.65533, 121.795 ])

If my above intuition is correct, then the corrective code to produce the right order of BTC prices from the shifted data frame is

bitcoin_prices_nbeats = bitcoin_prices_nbeats[bitcoin_prices_nbeats.columns[::-1]]

After executing the above line of code, X_train.iloc[0].to_numpy() outputs array([123.65499, 125.455 , 108.58483, 118.67466, 121.33866, 120.65533, 121.795 ])

The same is the case when loading full dataset to model_9. The corrective code for that corresponding dataset is X_all = X_all[:, ::-1]

Thanks in advance 😀 !!