how limit predict value between 0~1 or 1~100 #537
-
|
when i predict a value beyond the max value |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 2 replies
-
|
Hi @jaffe-fly For most applications, you could simply post-process the forecasts to cap them - would that be a viable option for you? |
Beta Was this translation helpful? Give feedback.
-
|
What I thought is that if it is greater than max, it is directly equal to max. I don't know if there is a better way |
Beta Was this translation helpful? Give feedback.
-
|
yeah, you could simply do something like df['yhat1'] = np.maximum(0, np.minimum(1, df['yhat1'].values)) |
Beta Was this translation helpful? Give feedback.
Hi @jaffe-fly
As of now, there is no option to limit, though logistic growth (which does more than just cap) may be added in the future.
For most applications, you could simply post-process the forecasts to cap them - would that be a viable option for you?