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

st: predict bug?

From   Stefan Kreisel <>
Subject   st: predict bug?
Date   Mon, 08 Jun 2009 17:37:13 +0200

Hi all.
When calculating a model with "ologit" (same bug? with "oprobit" or "regress" for that matter) with many independent categorical variables (25 in all; model has >2500 observations) and then using "predict yhat, xb" I get a bunch of cases (not all) with predictions=0; when calculating the linear predictor by hand, the results are perfectly identical to the stata output of the other non-zero predictions and the cases where stata predicted a 0, all calculate to non-zero numbers. Any ideas?
Stefan Kreisel

*   For searches and help try:

© Copyright 1996–2015 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index