[R] A question on glmnet analysis

[細田弘吉]

(11/03/26 0:12), Mike Marchywka wrote:
>
> Would you post your data or if you did I missed it?

Sorry, I did not. This is patients' data and I have to obtain some 
permission from my office to send it, so ...

> I don't have an answer for you but curious as I was
> interested in similar analyses with mostly continuous prognostic
> factors and continuous outcome ( survival). My interest
> was making a blind test to see if analysis could separate drug
> and treatment patients in a "triple blind" test.
> The end result would create preidctions ( drug or treeatment for
> each patient) and that would need to be compared to
> real assignments and checked for significance.

I want to divide my data into training and test data. But I only have 
104 observations. That's a problem.

Best regards,
KH