I have no idea. A data set that generates the error would be very helpful to me. What is the role of the last line BTW, the one with "1%" on it? Looking at the code I would guess that the vector "tied" has an NA in it, but how that would happen I can't see. There is a reasonable chance that it will become obvious once I have an example. Terry Therneau On 03/06/2015 05:00 AM, r-help-request at r-project.org wrote:> I don?t understand an error message from a thielsen function call within a > dplyr do function call. > > by.CaseNo <- group_by(as.data.frame(MAP.dt), CaseNo) >> >MAP.thielsen <- by.CaseNo %>% > + do(model = thielsen(noninvMAP ~ invMAP, symmetric = T, data = ., > + x = T, y = T, model = T, nboot = 1000)) > | | 1% ~40 m remaining > > > Error in if (any(tied)) { : missing value where TRUE/FALSE needed >
> I have no idea. A data set that generates the error would be very helpful to > me. What is the role of the last line BTW, the one with "1%" on it? > Looking at the code I would guess that the vector "tied" has an NA in it, but > how that would happen I can't see. There is a reasonable chance that it will > become obvious once I have an example.In a small bootstrap or other simulation one can end up by chance with a vector that doesn't behave; for example one with all values the same or with the one NA substituted into the whole vector. A common symptom is that the bootstrap sometimes works and sometimes doesn't, or only seems to work for small numbers of bootstrap replicates. But you're right; with no data, one can only guess. S Ellison ******************************************************************* This email and any attachments are confidential. Any use...{{dropped:8}}