singular matrix on training GMM model

[HongminWu]

Recently, I often encountered the singular matrix for training the GMM using your provided GMM class. However, the same data can be successfully trained. Would you have any time to check it out?

File "../models/gaussian.py", line 421, in pdf
return mvn.pdf(x, self.mean, self.cov, allow_singular=False) # default
File "/home/user/anaconda3/envs/py37/lib/python3.7/site-packages/scipy/stats/_multivariate.py", line 516, in pdf
psd = _PSD(cov, allow_singular=allow_singular)
File "/home/user/anaconda3/envs/py37/lib/python3.7/site-packages/scipy/stats/_multivariate.py", line 165, in init
raise np.linalg.LinAlgError('singular matrix')
numpy.linalg.LinAlgError: singular matrix

[lrozo]

Hi!
Well, there may be too many reasons why this is happening. However, the details you provided are not informative enough. It would be good to have a look at how you are using the class, how you initialize the GMM model (e.g., k-means, num of components, etc) and how you train it. Additionally, having a look at the data (plots may help) would be also nice to have a better understanding about the bug you are facing.
Not sure if you already solved this issue, but in case you have not, the above may help us to figure out what is going on.