This is the companion gitlab repository for our [arXiv manuscript](https://arxiv.org/abs/2102.12293)_"Two-way kernel matrix puncturing: towards resource-efficient PCA and spectral clustering"_ by Romain Couillet, Florent Chatelain and Nicolas Le Bihan
This is the companion gitlab repository for our [arXiv manuscript](https://arxiv.org/abs/2102.12293)_"Two-way kernel matrix puncturing: towards resource-efficient PCA and spectral clustering"_ by Romain Couillet, Florent Chatelain and Nicolas Le Bihan.
## Supplementary materials
...
...
@@ -11,8 +11,7 @@ The elements of proof of the main theorems in the core article are detailed in
### GAN Data
[VGG](https://arxiv.org/pdf/1409.1556.pdf) features of randomly [BigGAN](https://arxiv.org/abs/1809.11096)-generated images that we use in the paper are stored as [Git LFS](https://github.com/git-lfs/git-lfs/wiki) objects which avoids storing them locally. However this is possible to store the corresponding `.npy` files locally in order to avoid downloading them each time they are processed in the codes. This can be done manually via the GitLab web interface. An alternative, once Git LFS is [installed](https://github.com/git-lfs/git-lfs/wiki/Installation), this can be done with the command
[VGG](https://arxiv.org/pdf/1409.1556.pdf) features of randomly [BigGAN](https://arxiv.org/abs/1809.11096)-generated images that we use in the paper are stored as [Git LFS](https://github.com/git-lfs/git-lfs/wiki) objects which avoids storing them locally. However this is possible to store the corresponding `.npy` files locally in order to avoid downloading them each time they are processed in the codes. This can be done manually via the GitLab web interface. Alternatively, this can be done, once Git LFS is [installed](https://github.com/git-lfs/git-lfs/wiki/Installation), with the command:
