In case of a high dimensional data, if one wants to project it in a lower manifold using Diffusion Maps- what should be the values of the kernel sigma and no. of components ?
In R, using the package Destiny, it is easily possible to set the sigma to local and the number of components can be varied and different plots in different difusion components gives an idea typically which ones are rather relevant.
But in a python implementation, how does one make the sigma to be local (Can you suggest a good module). Also, if it remains global what is way to test which is the best sigma value. Finally, how do I make plots in different components ?
Himadri Nath Saha
http://scikit-learn.org/stable/auto_examples/decomposition/plot_pca_vs_fa_model_selection.html
0 votes
0 thanks
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Programming Languages
- Python
More Sabyasachi Dasgupta's questions See All
Similar questions and discussions