Closed Aneoshun closed 6 years ago
Hey @Aneoshun,
thanks for reporting. Let me check and get back to you.
I tried adding transpose in PR #46 and it fixed the problem indeed. Although the tests are failing on Travis, it might be due to Eigen versions.
Hi all,
Thanks for this library, I think it can be useful for my own research.
I am trying to use the projection function of the PCA. However, the result of the project is always a vector with the length of the high-dimensional space. Also, a dimension analysis of the matrix/vectors of the projection function, shows that there is something weird. Is this a bug or am I missing something?
I suspect that there is a .transpose() missing in the definition of the projection function Here
I tested this on my side, and I managed to retrieve the embedding after projecting the training dataset with my modified projection function.
Thanks