Re: [R] MDS, line of best fit, and id of variables

2016-10-18 Thread PIKAL Petr
Hi > -Original Message- > From: R-help [mailto:r-help-boun...@r-project.org] On Behalf Of Kirsten > Green > Sent: Tuesday, October 18, 2016 9:27 PM > To: David L Carlson > Cc: r-help@r-project.org > Subject: Re: [R] MDS, line of best fit, and id of variables > >

Re: [R] MDS, line of best fit, and id of variables

2016-10-18 Thread Bert Gunter
You need to spend time with an R tutorial or two. Your error is due to ignorance of basic R data structures. Bert On Oct 18, 2016 10:48 PM, "Kirsten Green" wrote: > David, > > I have run the str() function and all my data is int (integer). So I am > trying to change it to numeric using: > str(m

Re: [R] MDS, line of best fit, and id of variables

2016-10-18 Thread Kirsten Green
David, I have run the str() function and all my data is int (integer). So I am trying to change it to numeric using: str(mortdata) class(mortdata) is.numeric(mortdata) mortdata.num <- as.numeric(data.frame(mortdata)) But I keep getting an error: > mortdata.num <- as.numeric(data.frame( mortdata))

Re: [R] MDS, line of best fit, and id of variables

2016-10-18 Thread David L Carlson
0 stress 0.2838487 *** No convergence -- monoMDS stopping criteria: 20: stress ratio > sratmax --------- David L Carlson Department of Anthropology Texas A&M University College Station, TX 77840-4352 From: Kirsten Green [mailto:kagbo...@gmail.c

Re: [R] MDS, line of best fit, and id of variables

2016-10-18 Thread David L Carlson
What do you get with str(mortdata)? The error message indicates that at least one of the variables is not numeric and the second suggests it is a factor. You said the values were coded binomially, but they must be numeric, e.g. 0, 1 and not "Present" "Absent" or something like that. If they are