Thanks for this awesome repo, great learning
In Custom_ts_multi_data_prep function
X takes out indices (0, 47+1) and y takes indices (49, 59), rather y should take values from (48, 58)
range (0, 48) means it takes 0-47 indices since 48 is exclusive.
indicey = range(i+1, i+1+ horizon)
Is it supposed to be ???
indicey = range(i, i+horizon)
Using this improves R2 as welll. Correct me if am wrong ??
[link](https://github.com/Apress/hands-on-time-series-analylsis-python/blob/master/Chapter%203/2.%20Double%20Exponential%20Smoothing%20.ipynb) for the ipynb file I am talking about.
Here in the third cell there are some unnecessary (according to me)
df['Close'] = df['Close']
I dont think this was necessary yet it is included in the jupyter notebook.
Also very lengthy output is given for no specific reason.
In Custom_ts_multi_data_prep function for multivatrate time series analysis in chapter 7 need correction in every chapter discussed in that chapter.
X takes out indices (0, 47+1) and y takes indices (49, 59), rather y should take values from (48, 58)
range (0, 48) means it takes 0-47 indices since 48 is exclusive.