Hello everybody I am trying to extract some observations from a data frame, ie the subjects that belong to a given group, and although t-test etc work if I try to obtain the number of subjects in the subgroup i get some funny numbers. All the subjects with NA for group are included in the subgroups 'group==1' etc. Is this a bug? for example in Windoze R : Copyright 2001, The R Development Core Team Version 1.2.2 (2001-02-26)> z1 <- rnorm(100) > z2<- rep(1:5,20) > z3 <- z2 > z3[z1>1] <- NA > z1[z2==1][1] 1.16747148 0.58088013 1.00958356 1.35477371 - 1.19790468 0.07028495 -0.67707300 [8] -0.10999675 1.78765730 0.52044707 0.47965798 - 2.05768565 1.16349068 -0.75485570 [15] -0.70517830 0.86419410 -0.03514043 -0.84861275 - 0.50284191 0.64971596> z1[z3==1][1] NA 0.58088013 NA NA NA NA NA [8] -1.19790468 NA 0.07028495 -0.67707300 -0.10999675 NA NA [15] NA 0.52044707 0.47965798 -2.05768565 NA - 0.75485570 NA [22] -0.70517830 0.86419410 NA -0.03514043 -0.84861275 NA -0.50284191 [29] NA NA NA 0.64971596>the first with z2==1 gives me the 20 subjects OK but the second should have given me less than 20 but it returns 32! it adds all the NA's for z3. I would have thought that z3==1 excluded all the NA's Thanks for any explanation. R. Heberto Ghezzo Ph.D. Meakins-Christie Labs McGill University Montreal - Canada heberto at meakins.lan.mcgill.ca -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
On Thu, 19 Apr 2001, Heberto Ghezzo wrote:> Hello everybody > I am trying to extract some observations from a data frame, ie the > subjects that belong to a given group, and although t-test etc work > if I try to obtain the number of subjects in the subgroup i get some > funny numbers. All the subjects with NA for group are included in > the subgroups 'group==1' etc. > Is this a bug?The description of BUGS in the FAQ will help you answer your own question.> for example in Windoze???> R : Copyright 2001, The R Development Core Team > Version 1.2.2 (2001-02-26) > > > > z1 <- rnorm(100) > > z2<- rep(1:5,20) > > z3 <- z2 > > z3[z1>1] <- NA > > z1[z2==1] > [1] 1.16747148 0.58088013 1.00958356 1.35477371 - > 1.19790468 0.07028495 -0.67707300 > [8] -0.10999675 1.78765730 0.52044707 0.47965798 - > 2.05768565 1.16349068 -0.75485570 > [15] -0.70517830 0.86419410 -0.03514043 -0.84861275 - > 0.50284191 0.64971596 > > z1[z3==1] > [1] NA 0.58088013 NA NA NA NA > NA > [8] -1.19790468 NA 0.07028495 -0.67707300 -0.10999675 > NA NA > [15] NA 0.52044707 0.47965798 -2.05768565 NA - > 0.75485570 NA > [22] -0.70517830 0.86419410 NA -0.03514043 -0.84861275 > NA -0.50284191 > [29] NA NA NA 0.64971596 > > > > the first with z2==1 gives me the 20 subjects OK but the second > should have given me less than 20 but it returns 32! it adds all the > NA's for z3. > I would have thought that z3==1 excluded all the NA's > Thanks for any explanation.NA means unknown. So you don't know if those values are one, and all you can get is a report that they might be. I suspect you intended z1[z3 %in% 1] -- Brian D. Ripley, ripley at stats.ox.ac.uk Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/ University of Oxford, Tel: +44 1865 272861 (self) 1 South Parks Road, +44 1865 272860 (secr) Oxford OX1 3TG, UK Fax: +44 1865 272595 -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
Hello. If I understood BDR correctly then when I want so subset a data frame, matrix etc. instead of doing x1 <- x[group == 1] i should do x1 <- x[group %in% 1] the second way being the preferred way to avoid problems with NA's? As always thanks to BDR for his prompt , and to the point, response. R. Heberto Ghezzo Ph.D. Meakins-Christie Labs McGill University Montreal - Canada heberto at meakins.lan.mcgill.ca -.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.- r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-help-request at stat.math.ethz.ch _._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._