On 06/09/2013 5:29 AM, Markus Gschwind wrote:
Hello,
I am a beginner of R but knowing SPSS and matlab.
I need to analyse my data with a mixed type cluster analysis, that's why I
am looking into R.
I have a datasheet with 45 subjects (rows) and 30 variables of each subjet
(columns) in Excel (datasheet.csv) or SPSS (datasheet.sav).
In SPSS I can define if the variable is ordinal (e.g. low - intermediate -
high), nominal (e.g. type of hobby) or scale (e.g. age).
The variables are all numerical (as SPSS needs them in numbers coding the
ordinal or nominal qualities).
However, when using data importing like
mydata<-read.spss("datasheet.sav")
or
mydata<-read.table("datasheet.csv", sep=",",na="",stringsAsFactors=F)
this produces a 'list' and not a 'data frame' with factors.
My question: How can I import the "measure" (i.e. nominal, ordinal, scale)
form SPSS?
Or how can I define variables in mydata as nominal or ordinal, even if they
consist of numbers?
Actually, the read.table() function will produce a dataframe, but you
asked for no factors, so you'll get no factors.
To convert a variable x to a nominal factor, use
f <- factor(x, levels = c( ..... ))
where the levels are chosen to list all possible values that x could
take. If x is a column of a dataframe d, use
d$f <- factor(d$x, levels = c( .... ))
to create a new column named f in the dataframe.
For an ordinal factor, do the same, but use ordered() instead of factor().
(You can do some of this automatically using read.table, e.g. by
specifying colClasses, but you have more flexibility if you do it as
shown above.)
Duncan Murdoch
______________________________________________
R-help@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.
