Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: bootstrap validation and c statistics (help me please !)

From   Nick Cox <[email protected]>
To   [email protected]
Subject   Re: st: bootstrap validation and c statistics (help me please !)
Date   Wed, 25 May 2011 07:56:21 +0100

Statalist is a discussion list, not a help line that guarantees
support to anyone who asks for it. You can post a question, but
someone else must want to discuss it. Of your three questions, the
first two are too vague or too general for me even to know what kind
of answer is appropriate, although no one else is obliged to agree.
But in general I think it's in your own interest to ask questions that
are likely to be answered.

The third fails to satisfy standard advice on this list, which is

"Say exactly what you typed and exactly what Stata typed (or did) in
response. N.B. exactly! If you can, reproduce the error with one of
Stata's provided datasets or a simple concocted dataset that you
include in your posting."

The report of "an error" is on fours with reporting that a car or a
television is not working; any support person is expected to show
sympathy, but needs more information.  What precisely is your error?


On Tue, May 24, 2011 at 11:38 PM, Hong, Sounman
<[email protected]> wrote:

> I need to validate my model. I heard that I need to do either by a cross-validation or a bootstrap. I have three questions.
> 1. What is the best statistics to report if I want to show that my model is validated using bootstrap? I read a paper that reported C statistics, although I have to study what that is. Do you recommend me reporting C statistics?
> 2. Does STATA have a code that generates a shrinkage factor?
> 3. I wrote below my STATA code that performs bootstrap and get a C statistics. But I got an error and I don't know why. Please advise.
> use bootstrap.dta, clear
> logit DEPVAR BPA
> estat concordance
> matrix b = e(b)
> capture program drop mysim
> program define mysim, rclass
>    use bootstrap.dta, clear
>    bsample
>    lroc BPA, nograph beta(b)
>    return scalar c=r(c)
> end
> simulate c=r(c), reps(100): mysim
> Thank you in advance. Any advice would be greatly appreciated.

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index