Open mikera opened 10 years ago
I was not aware of that library, thanks for dropping note, core.matrix seems a really great library! Generally, I did not want to use a matrix library as a dependency because all of the computation other than training data and test data is specifically is done 1-D vector operations. That being said, I wanted do provide some util functions based on pairwise distance matrix if one first wants to reduce dimensionality first(like PCA or maybe MDS), core.matrix seems perfect for that.
If you have a pull request, feel free, they are more than welcome!
Hi there, looks like a really promising library!
Have you consider implementing the algorithms in terms of core.matrix (https://github.com/mikera/core.matrix)?
This would give a few advantages: