Thanks as always to Kit Baum, a new package srslogit is now available for download from SSC. In Stata, use the ssc command to do this.
The srslogit package is described as below on my website. It fits a primary logit model and (optionally) a secondary ridit spline model in the ridit of the predicted odds. It requires the existing SSC packages wridit and bspline in order to work, and optionally adds to the dataset new variables, containing the predicted odds, probabilities, and inverse probability weights for the 2 logistic models. Its main application is probably in computing treatment-propensity weights and completeness-propensity weights.
Best wishes
Roger
---------------------------------------------------------------------------------
package srslogit from http://www.rogernewsonresources.org.uk/stata16
---------------------------------------------------------------------------------
TITLE
srslogit: Logit regression with secondary ridit splines
DESCRIPTION/AUTHOR(S)
srslogit fits a primary logit model for a binary dependent variable
in a list of independent variables, followed optionally by a
secondary ridit spline model for the same binary dependent variable
in the ridit of the predicted dependent variable from the primary
logit model. It optionally outputs variables created in this
process, including the predicted values (also known as propensity
scores) and inverse probability weights (also known as inverse
propensity weights) from the primary and secondary logit models.
These propensity weights may be used as treatment propensity weights
(if the dependent variable is a treatment indicator) or as
completeness-propensity weights (if the dependent variable is a
completeness indicator). srslogit uses the SSC packages wridit and
bspline, which have to be installed in order for srslogit to run.
Author: Roger Newson
Distribution-Date: 28august2021
Stata-Version: 16
INSTALLATION FILES (click here to install)
srslogit.ado
srslogit.sthlp
---------------------------------------------------------------------------------
(click here to return to the previous screen)
0 Response to New package srslogit on SSC
Post a Comment