Multiplying transposed matrices in mir

jmh530 john.michael.hall at
Mon Apr 20 19:12:52 UTC 2020

On Monday, 20 April 2020 at 19:06:53 UTC, p.shkadzko wrote:
> [snip]
> It is. I was trying to calculate the covariance matrix of some 
> dataset X which would be XX^T.

Incorrect. The covariance matrix is calculated with matrix 
multiplication, not element-wise  multiplication. For instance, I 
often work with time series data that is TXN where T > N. 
Couldn't do a calculation with element-wise multiplication in 
that case.

Try using Lubeck's covariance function or checking your results 
with the covariance function in other languages.

More information about the Digitalmars-d-learn mailing list