in R, I perform Full Matching with different specifications.
This means that I use different distances (glm, randomforest, cbps), discard variables (none, both, control, treated) and calipers (none, 0.1 and 0.2).
After that, I need to extract the Match Matrix or otherwise observe the matching progress, but I have no option to do so. For example, if I run:
m.out <- matchit(formula = formula, data = data, method = "full", discard = "control", distance = "glm", link = "logit")
,than I am not able to run this:
m.out$match.matrix
in order to observe the matched pairs.
Is there anything obvious what I am missing? I first started to doubt whether there are matches at all (and this is subclassification only), but it is called "Full MATCHING" and the definitions for this algorithms do not contradict to this fact, as far as I understand it.
I tried every combination of calipers, discards, and distances, but there are no Match Matrizes to observe.
One explanation I found was that "matching strata are not indexed by treated units as they are in some other forms of matching" (https://kosukeimai.github.io/MatchIt/reference/method_full.html), but is there another way how to assess the matched pairs with MatchIt?
I would really appreaciate some help and clarification in there.
Thank you!