hmmm.... It's hard to tell just from this. Can you replicate this error in, say, the Lalonde data set? Or is this an error specific to your data and application? Kosuke -- Department of Politics Princeton University http://imai.princeton.edu On Wed, 1 Apr 2009, Beth Gifford wrote: Dear colleagues, > > I am using MatchIt for my project. Cfst_std is a treatment variable. We > want > to use ?exact? matching in gender and use ?nearest? matching for other > variables. However, I got the error message: Error in > weights.matrix(match.matrix, treat, discarded). What does this mean? And > how > do I fix this issue? > > > > Thanks so much, > > > > cfst_std <-yu$cfst_std >> > > table (cfst_std) >> > > cfst_std > > 0=no 1=yes > > 25034 3074 > > > > summary (yu$boy) >> > > 0=no 1=yes > > 14042 14066 > > >> > m.out <- matchit(cfst_std~math_s+old_grade, data=yu, exact=c("boy"), >> > method="nearest") > > Error in weights.matrix(match.matrix, treat, discarded) : > > No units were matched > > In addition: Warning messages: > > 1: In max(pscore[treat == 0]) : > > no non-missing arguments to max; returning -Inf > > 2: In max(pscore[treat == 1]) : > > no non-missing arguments to max; returning -Inf > > 3: In min(pscore[treat == 0]) : > > no non-missing arguments to min; returning Inf > > 4: In min(pscore[treat == 1]) : > > no non-missing arguments to min; returning Inf > > > Yu > >