[R] svm plot
Δέσποινα Καλφακάκου
dee88 at windowslive.com
Fri Mar 13 12:38:26 CET 2015
Hello,
as you can imagine, I am new to R. I have a problem plotting the results of my svm classification of microarray data.
So, I have a matrix of 30 rows (samples) and 306 columns (significant genes). I transform the matrix to a data frame in order to do the classification.
My code is:
data<-t(significantGenes)
labels<-gl(2,20,30)
data<-data.frame(data)
svm.cross <- svm(data, labels, type="C-classification", kernel="polynomial", degree=5, gamma=0.5, coef0=1, cross=10)
This code works great!
But when I try to plot the results (using many different ways listed online), I get the following error:
"Error in terms.default(x) : no terms component nor attribute"
Here is SOME of the ways I've tried to plot the result:
plot(svm.cross, data, formula = V1~V2)
plot(svm.cross, data, formula = V1~V2, slice = list(V1=2,V2=2))
plot(svm.cross, data, formula = data[,1]~data[,2])
and
plot(svm.cross, data,formula=data["V1"]~data[V2]) with error: "Error in model.frame.default(formula, data) :
invalid type (list) for variable 'data["V1"]'"
Note that str(data) gives: 'data.frame': 30 obs. of 306 variables:
$ V1 : num 10.8 10.3 12.5 12.7 11.7 ...
$ V2 : num 9.5 10.34 11.25 11.56 9.74 ...
$ V3 : num 9.36 11.67 10.52 11.57 10.25 ...
$ V4 : num 9.25 8.52 10.61 11.21 9.84 ...
etc.
Thank you so much for your response.
Kind regards,
Despina
[[alternative HTML version deleted]]
More information about the R-help
mailing list