1

How to get Standard Deviation from each components in sklearn GMM after fit?

model.fit(dataSet)
model.means_ is the means of each components.
model.weights_ is the co-efficient of each components.

Where I can find the standard deviations of each Gaussian components?

Arun
  • 21
  • 1
  • 2
  • You can get the covariance matrix using `model.covariances_` which contains the information that you're after. – piman314 Apr 10 '18 at 09:48
  • Can you able to elaborate an bit more ? – Arun Apr 10 '18 at 10:33
  • It depends on the covariance structure you've used in your model and exactly what you're trying to do. If you can update your questions with a MWE and a bit more detail I can try to help. – piman314 Apr 10 '18 at 10:36
  • A table of random values (sampled from two Gaussian distributions) is provided in: https://drive.google.com/openid=1LIGIpq96Bkg_z6TGtEaTv0y1Q2sjipZ4 What are the standard deviations of the two Gaussians from which the sample data is drawn? – Arun Apr 10 '18 at 11:09
  • That link is dead. Also can you make sure that you edit your question, it makes the question and answer much easier to read for other. Sorry I know this is a pain, but if your Gaussians are 10 dimensional then standard deviation doesn't really mean anything, so it's best to be super clear and give code to either generate dummy data or a link to actual data. – piman314 Apr 10 '18 at 15:38
  • updated link: https://drive.google.com/drive/folders/1e-mwPZwjQ4QgfNV8H5A6-lSH7nLoFN8q?usp=sharing – Arun Apr 10 '18 at 16:11
  • 1
    Your data in 1 dimensional, so `model.covariances_` will just return the variance of each component. – piman314 Apr 10 '18 at 16:31
  • Thank you , Is it possible to find the relative proportion of these two mixture ? – Arun Apr 10 '18 at 16:45

0 Answers0