Hello everyone! Is anyone familiar with a function/package/implementation of the MDS INDSCAL (individual differences scaling) algorithm for Python? Any advice would be super appreciated. Thanks! 
Where or how have you looked?
By searching for INDSCAL/individual differences scaling/MDS (and combinations thereof) on Google, checking out various forums, packages documentation and talking to colleagues that use Python or used MDS before, etc. There are implementations in SPSS and R, but all the other scripts for analyses we have for this project are in Python, so it’d be good to keep it consistent. There is, of course, MDS module in Sklearn, but as far as I understand, this does not include the option for individual differences scaling. I’m not familiar with all the packages available for Python of course so I might be missing something obvious, that’s why I thought I’d ask experts.
OMG, thank you so much! I don’t know how I missed this! Thanks!
Hi Magdalena,
Have you managed to use INDSCAL from sktensor?
I am using one of the forks (mostly to have updated dependencies).
However, the API is a bit cryptic (indscal.py):
def orth_als(X, ncomp, **kwargs)
def init(X, init, ncomp)
No hints about input format nor about what each function is intended for.
The MDS implementation in scikit learn can be found in the sklearn.manifold module. To use MDS with INDSCAL, you will need to prepare your dissimilarity or similarity data and create an instance of the MDS class. Then, you can fit the MDS model to your data using the fit_transform method.