negative divergence value

[vaibhav0195]

hi thanks for the repository... I was using the code to calculate the divergence between gaussian distribution but I was getting negative divergences. Mean vectors for both gaussians was a 2 dimensional vector, each entry was chosen randomly between (0,1) and covariance matrix was an identity matrix. Calculating Kl divergence by using K = NONE gives negative values.

[vaibhav0195]

update : is i use K =1 or 5 I am still getting negative values. I thought KL divergence can not be negative. Is there something I am missing ?? any help is appreciated. thanks