在R中使用mapply运行逻辑模型

Question

I'm having the following function:

logModels <- function(data_idx_list,data)   

{
  x  <-  lapply(data_idx_list, function(m) 
        sparse.model.matrix(~.,data = data[m,]))
  y <- lapply(data_idx_list, function(m) data$earlyR[m])

  logM <- mapply(function(x,y) {
          cv.glmnet(x=x,y=y,family="binomial",alpha=0)
}, x,y) 

return(logM)

}

where x contains 5 samples of data, and y is my dependent variable . When I'm trying to use mapply I'm getting the following error:

Error in glmnet(x, y, weights = weights, offset = offset, lambda = lambda, : number of observations in y (1) not equal to the number of rows of x (100000)

But, when I'm running the model as follows it works:

lm1 = cv.glmnet(x=x[[1]],y=y,family="binomial",alpha=0)

So, I assume I have some trouble with the way I'm approaching x in my mapply function.

Your help will be appreciated.

Answer 1

If I understand your problem correctly, you simply need to wrap y into a list :

logM <- mapply(function(x, y) cv.glmnet(x, y, family = "binomial", alpha = 0),
               x, list(y))

Otherwise individual elements of y are passed to consecutive invocations of your glm function.

Another, more concise way of writing this:

logM <- mapply(cv.glmnet, x = x, y = list(y), family = "binomial", alpha = 0))