The answer is simple data[,4] == 1 produces a logical vector of length nrow(data) and the subsetting mechanism for data frames in R needs a vector of the same length as the data frame has rows.
data[1:20,4] == 1 produces a data frame of length 20, and if this is not the length of data. So R applies its standard procedure, it repeats this vector as often as needed to get a vector of length == nrow(data) Th following code illustrates what is happening data <- data.frame(x=rnorm(100),y=rnorm(100),z=rnorm(100),a=rep(c(1,2,1,2),c(2,48,2,48))) vec1 <- data[,4]==1 vec2 <- data[1:20,4]==1 > On 27 Feb 2017, at 13:07, Tunga Kantarcı <tungakanta...@gmail.com> wrote: > > Consider a data frame named data. data contains 4 columns and 1000 > rows. Say the aim is to bring together columns 1, 2, and 4, if the > values in column 4 is equal to 1. We could use the syntax > > data(data[,4] == 1, c(1 2 4)) > > for this purpose. Suppose now that the aim is to bring together > columns 1, 2, and 4, if the values in column 4 is equal to 1, for the > first 20 rows of column 4. We could use the syntax > > data(data[1:20,4] == 1, c(1 2 4)) > > for this purpose. However, this does not produce the desired result. > This is surprising at least for someone coming from MATLAB because > MATLAB produces what is desired. > > Question 1: The code makes sense but why does it not produce what we > expect it to produce? > > Question 2: What code is instead suitable? > > ______________________________________________ > R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see > https://stat.ethz.ch/mailman/listinfo/r-help > PLEASE do read the posting guide http://www.R-project.org/posting-guide.html > and provide commented, minimal, self-contained, reproducible code. ______________________________________________ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
