问题
3 days old to R and can't figure out what I'm doing wrong. I'm trying to send some columns with two way interactions into a glmnet cox model. I have some data.frame() called dtable
Edit to make the code reproducible
xs<-c("Col1", "Col2", "Col3")
v<-c(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, NA, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, NA, 25, 26, 27, 28, 29, 30)
df<-data.frame(matrix(v,ncol=3))
dm<-as.matrix(df)
dm<-matrix(dm[complete.cases(dm)], ncol=3)
colnames(dm)<-xs
dfdata<-data.frame(dm)
f<-as.formula(time~.*.)
xmatrix<-model.matrix(f, dfdata)[,-1]
When I run this I get the error
Error in model.frame.default(object, data, xlev = xlev) :
object is not a matrix
Thanks in advance
回答1:
It's because of the formula: time~.*. In the data.frame() there is no time column for the formula to cross everything by.
回答2:
I suspect the error is telling exactly what is wrong. The object you are passing to model.matrix()
is not a matrix. What is the result of class(data)
? Probably a data.frame.
Try adding as.matrix()
to the model.matrix()
call to data
.
Two other notes - don't call your data.frames data
. Also asks questions here using a reproducible example. You'll get better responses.
回答3:
I think instead of this:
dm<-matrix(dm[complete.cases(dm)], ncol=3)
you want this:
dm <- as.matrix(dm[complete.cases(dm), ])
Not to toot my own horn (too much), but consider using glmnetUtils to simplify the task of managing data frames, formulas and model matrices.
来源:https://stackoverflow.com/questions/40682261/error-in-model-frame-defaultobject-data-xlev-xlev-object-is-not-a-matrix