Open tejendrapsingh opened 4 years ago
Hi did you fix this issue? I ran into the exact error with you and I have no idea how to solve this, too.
I think are no valid matches for some subgroups.
On Wed, Jul 20, 2022 at 5:43 PM cocoxxo @.***> wrote:
Hi did you fix this issue? I ran into the exact error with you and I have no idea how to solve this, too.
— Reply to this email directly, view it on GitHub https://github.com/JasjeetSekhon/Matching/issues/2#issuecomment-1190787131, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADHXHINSTWJ4MM3WVZ3HU7DVVBXJBANCNFSM4LDNUDRQ . You are receiving this because you are subscribed to this thread.Message ID: @.***>
Did you remove the subgroup in order to do this?
On Wed, Jul 20, 2022 at 19:13 Jasjeet Sekhon @.***> wrote:
I think are no valid matches for some subgroups.
On Wed, Jul 20, 2022 at 5:43 PM cocoxxo @.***> wrote:
Hi did you fix this issue? I ran into the exact error with you and I have no idea how to solve this, too.
— Reply to this email directly, view it on GitHub < https://github.com/JasjeetSekhon/Matching/issues/2#issuecomment-1190787131 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/ADHXHINSTWJ4MM3WVZ3HU7DVVBXJBANCNFSM4LDNUDRQ
. You are receiving this because you are subscribed to this thread.Message ID: @.***>
— Reply to this email directly, view it on GitHub https://github.com/JasjeetSekhon/Matching/issues/2#issuecomment-1190857408, or unsubscribe https://github.com/notifications/unsubscribe-auth/ANMRVMYC5RT4567243KILVDVVCBZDANCNFSM4LDNUDRQ . You are receiving this because you commented.Message ID: @.***>
Looks like the GenMatch function called by the Matchby function has some issues. I am posting the code and the associated error below. Thanks.
Code:
set.seed(12345)
rr <- Matching::Matchby(#Y = Y,
Tr = Tr, X = my.X, by = by1, M = 1, replace = TRUE, estimand = "ATT", ties = FALSE)
Error Message:
60 of 5106 groups
Error in if (Sig.X[k, 1] < tolerance) Sig.X[k, 1] <- tolerance : missing value where TRUE/FALSE needed In addition: Warning message: In sqrt(max(tolerance, sum(X[, k] * X[, k] * weights))/weights.sum - : NaNs produced