0

Here's a simplified explanation: I have a dataframe of multiple unrolled time-steps that is used to predict price2 and volume2 (the next step). Before I train my network, I want to use StandardScaler.

However, when I went to invert the predicted that data, I now have the challenge that the StandardScaler is adjusted to lenght of 4 and hence cannot be applied to my output length of 2. 

df = pd.DataFrame({'price':10, 'price1':20, 'volume':100, 'volume1':75})

What is the best way to deal with this?

Jarad
  • 17,409
  • 19
  • 95
  • 154
johan bender
  • 53
  • 1
  • 1
  • 5
  • StandardScaler is used only for the normalization of the data. What do you want to do exactly? Can you add the full code and data ? – seralouk Jul 24 '18 at 17:42
  • Are you sure you are standardizing the way you wish? StandardScaler will scale per column by default, but it sounds like you want it to scale per series. – Bert Kellerman Jul 27 '18 at 19:43

0 Answers0