tslearn-team/tslearn

Scalers inverse_trasnsform() function

Lingkangjin opened this issue · 2 comments

Problem
Is it possibile to transform back after the machine learning process,
for instance I did the preprocessing of my data using "TimeSeriesScalerMeanVariance()" and did k-means clustering, therefore I have found the clusters centroids, however I could not to get back to original data, i.e. inverse of "TimeSeriesScalerMeanVariance()", is there solution to this?

Describe the solution you'd like
I would like to have something like Scikitlearn based method, which allows to inverse back directly

Dear @Lingkangjin

There is no such thing as a general inverse transform in this case, since scaling is done per series, as mentionned in the docs.

Dear @rtavenar,

Thanks for the prompt response,
Well noted