From: Terry Lin on
When using the function coxphfit, how do I handle the data input when each of the predictors (covariates) has different vector length? Many thanks
From: Peter Perkins on
On 6/23/2010 9:20 PM, Terry Lin wrote:
> When using the function coxphfit, how do I handle the data input when
> each of the predictors (covariates) has different vector length? Many

Terry, you're going to have to provide more information. The matrix of
covariates that you provide as an input to COXPHFIT should represent n
observations on p variables, and so by definition each covariate has the
same length. It may be that you have some missing values. That's
ordinarily handled by inserting NaNs in those elements, and COXPHFIT
ignores any row with a NaN in it.
From: Georgina Hunter on
Hello Can Terry Lin please contact me. Thanks
 | 
Pages: 1
Prev: save workspace AND figures
Next: FMINCON Message