1

I am using PyEMD (Empirical Mode Decomposition) on a signal (train & test ) data. All seems be working fine for all signals (datasets), but in my one of the dataset number of IMFs it is decomposing is different for train & test dataset.

I have tried (max_imf: ) argument, but by limiting the number to minimum value so that both (train & test ) have same number of IMF, the decomposition is not correct ( it is not decomposing till final trend).

Any suggest will be appreciated.

Mohd Naved
  • 358
  • 3
  • 12

1 Answers1

0

You can decompose the data first, and then divide the training set and the test set for each component

10 Rep
  • 2,217
  • 7
  • 19
  • 33
iris486
  • 1
  • 1