Dear list
I am using matchit to perform optimal matching, and would like to use
the discard="both" option to remove both rrated and control cases that
lie outside the option of common support. Is this possible with
method="optimal"? I have tried to employ discard="both" with full
matching (method="full") as well but I keep getting the same error:
Error in fullmatch.matrix(d, ...) : argument 'x' must have dimnames
In addition: Warning messages:
1: glm.fit: algorithm did not converge
2: glm.fit: fitted probabilities numerically 0 or 1 occurred
3: In fullmatch(d, ...) :
Without 'data' argument the order of the match is not guaranteed
to be the same as your original data.
The above errors are not presented when the discard option is omitted
from my code. Here is how I am invoking matchit:
matchit(v0obese ~ age + educ.hs + educ.scol + educ.bach + educ.sgs +
educ.gs +
race.black + race.other + v0copd + v0hrtfa + v0sbp + v0vitmnd + v0calcum +
v0bmi + v0ces_d + v0pase + v0_stepmi + v0lldiir + v0_wspami + v0_wspbmi +
v0_wspcmi + v0copd, method="full",data=most,discard="both");
I suspect that the discard procedure is causing some variables to have a
constant value, but I'm not sure.
Barth