I would like to do the following:
1- simulate 1000 observations
2 - simulate a continuous biomarker which can be normally distributed with a given mean and given standard deviation and simulate a binary outcome, and ensure the following:
a] the ROC AUC is 0.85
b] when fitting a logistic regression on the association between the predictor continuous biomarker on the binary outcome, I obtain an odds ratio of 1.12 (which would mean that for 1 unit increase of the biomarker, the odds of the event increase by 12%)
How do I do this in STATA or R?
Thanks
A
Related Posts with How to simulate data with specified ROC AUC and specified odds ratio
Goodness of fit for panel dataHello, Statalist community! I feel a little foolish posting this question but I'm hoping that someo…
Counting number of specific observations per groupDear all, I have a simple problem for which I have not been able to find a solution. Imagine a very…
estimate store for xtprobithello guys, I am using stata12 and I have a question regarding the estimate store for a xtlogit mod…
Building pretty crosstabs in putdocxHi Statalist- We are seeking some help in formatting tables in putdocx to look, well, pretty. We us…
Creating a subsequent variable after "max" commandHi, I'm having difficulty in creating a variable after using "max" command. Below illustrates it. …
Subscribe to:
Post Comments (Atom)
0 Response to How to simulate data with specified ROC AUC and specified odds ratio
Post a Comment