mds._smacof_single: tidy up and implementation details #11381
Labels
Comments
|
Despite the names of the variables the equations are coherent with theoretical work, nicely summarized in the book: |
|
#18933 makes a summary of issues with smacof algorithm. I'm closing this one as it is more a question than an issue. Feel free to open a new github discussion if more clarifications are needed. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Hi everyone,
I'm a bit puzzled about implementation of the SMACOF algorithm.
Especially in LL98-108 names of the variable a very confusing:
disparities,similarities,sim,disanddisparitiesare used interchangeably, which make the code difficult to understand. This could be tided up.I have several questions about implementation:
Why
disparitiesare multiplied bynp.sqrt((n_samples * (n_samples - 1) / 2) / (disparities ** 2).sum())(L107)?Could you please point on which theoretical derivations this implementation is based?
Maybe @NelleV can help?
The text was updated successfully, but these errors were encountered: