I need to create matched test and / control pairs from a (dis)similarity
matrix. I am not sure if the existing matching softwares (e.g., optmatch,
matchit ) will allow me to input my own dissimilarity matrix.
It seems that these algorithms are focusing more on observational studies
with the distance measure being operationalized by the probability of
belonging to the existing test and control cells.
I would appreciate any help here.
I guess I could otherwise simply look up the dissimilarity matrix to select
pairs. That kind of gets cumbersome if you have 65 by 65 matrix.