hi everyone, wondering if you could help me with a novice problem. I have a data frame called subjects with a height and weight variable and want to calculate a bmi variable from the two. i have tried: attach(subjects) bmi <- (weight)/((height/100)^2) but it comes up with the error: Warning messages: 1: In Ops.factor(height, 100) : / not meaningful for factors 2: In Ops.factor((weight), ((height/100)^2)) : / not meaningful for factors I presume that this means the vectors height and weight are not in numeric form (confirmed by is.numeric) so i changed the code to: bmi <- (as.numeric(weight))/((as.numeric(height)/100)^2) but this just comes up with a result which doesnt make sense i.e. numbers such as 40000 within bmi vector. Ive looked at as.numeric(height)/as.numeric(weight) and these numbers just arnt the same as height/weight which is the reason for the incorrect bmi. Cant anyone tell me where I am going wrong? Its quiet frustrating because I cant understand why a function claiming to convert to numeric would come up with such a bizarre result. -- View this message in context: http://www.nabble.com/help-with-as.numeric-tp23558326p23558326.html Sent from the R help mailing list archive at Nabble.com.
as.numeric() doesn't convert factors to the explicit value, nor should it. Under what you're expecting, ff you have a factor where the levels are "Female" and "Male", using as.numeric() wouldn't produce anything meaningful. However, as.numeric() does something much smarter. It converts "Female" to 1, and "Male" to 2. More generally, if you have n levels, it will produce a vector of values between 1 and n. This is referred to as the 'internal coding.' If you want to convert your height and bmi variables to their numeric values, you need to do> as.numeric(as.character(height))This will get you around the internal coding. -----Original Message----- From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org] On Behalf Of deanj2k Sent: Friday, May 15, 2009 7:58 AM To: r-help at r-project.org Subject: [R] help with as.numeric hi everyone, wondering if you could help me with a novice problem. I have a data frame called subjects with a height and weight variable and want to calculate a bmi variable from the two. i have tried: attach(subjects) bmi <- (weight)/((height/100)^2) but it comes up with the error: Warning messages: 1: In Ops.factor(height, 100) : / not meaningful for factors 2: In Ops.factor((weight), ((height/100)^2)) : / not meaningful for factors I presume that this means the vectors height and weight are not in numeric form (confirmed by is.numeric) so i changed the code to: bmi <- (as.numeric(weight))/((as.numeric(height)/100)^2) but this just comes up with a result which doesnt make sense i.e. numbers such as 40000 within bmi vector. Ive looked at as.numeric(height)/as.numeric(weight) and these numbers just arnt the same as height/weight which is the reason for the incorrect bmi. Cant anyone tell me where I am going wrong? Its quiet frustrating because I cant understand why a function claiming to convert to numeric would come up with such a bizarre result. -- View this message in context: http://www.nabble.com/help-with-as.numeric-tp23558326p23558326.html Sent from the R help mailing list archive at Nabble.com. ______________________________________________ R-help at r-project.org mailing list 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. ================================== P Please consider the environment before printing this e-mail Cleveland Clinic is ranked one of the top hospitals in America by U.S. News & World Report (2008). Visit us online at http://www.clevelandclinic.org for a complete listing of our services, staff and locations. Confidentiality Note: This message is intended for use\...{{dropped:13}}
On May 15, 2009, at 6:57 AM, deanj2k wrote:> > hi everyone, wondering if you could help me with a novice problem. > I have a > data frame called subjects with a height and weight variable and > want to > calculate a bmi variable from the two. i have tried: > > attach(subjects) > bmi <- (weight)/((height/100)^2) > > but it comes up with the error: > Warning messages: > 1: In Ops.factor(height, 100) : / not meaningful for factors > 2: In Ops.factor((weight), ((height/100)^2)) : > / not meaningful for factors > > I presume that this means the vectors height and weight are not in > numeric > form (confirmed by is.numeric) so i changed the code to: > > bmi <- (as.numeric(weight))/((as.numeric(height)/100)^2) > > but this just comes up with a result which doesnt make sense i.e. > numbers > such as 40000 within bmi vector. Ive looked at > as.numeric(height)/as.numeric(weight) and these numbers just arnt > the same > as height/weight which is the reason for the incorrect bmi. Cant > anyone > tell me where I am going wrong? Its quiet frustrating because I cant > understand why a function claiming to convert to numeric would come > up with > such a bizarre result.That 'height' is a factor suggests that you imported the data using one of the read.table() family of functions and that there are non- numeric characters in at least one of the entries in that column. Since 'height' is a factor, if you use as.numeric(), you will get numeric values returned that are the factor level numeric codes and not the expected numeric values. That is why you are getting bad values for BMI. See: http://cran.r-project.org/doc/FAQ/R-FAQ.html#How-do-I-convert-factors-to-numeric_003f If you use something like: grep("[^0-9\\.]", height, value = TRUE) that should show you where you have non-numeric values in the 'height' column. That is, entries for 'height' that contain characters other than numeric or a decimal. Foe example: height <- factor(c(seq(0, 1, 0.1), "1,10", letters[1:5])) > height [1] 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1,10 a b c d e Levels: 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1,10 a b c d e > as.numeric(height) [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 > grep("[^0-9\\.]", height, value = TRUE) [1] "1,10" "a" "b" "c" "d" "e" I would also check the 'weight' column for the same reasons, to be sure that you don't have bad data there. Another approach would be to use: str(subjects) which will give you a sense of the data types for each column in your data frame. Review each column and take note of any columns that should be numeric, but are factors. See ?str, ?grep and ?regex for more information. You might also want to look at ?type.convert, which is the function used by the read.table() family of functions to determine the data types for each column during import. HTH, Marc Schwartz