NEWS

---

saws 0.9-7.0 (2022-06-23)

• the geeUOmega function had an error, and it gave the wrong answer when there was an offset. Thanks to Amy MacDougall for pointing this out.
• the print.saws did not print out the test matrix and beta0 if beta0 NE 0. Now it does. This is explained in the man page for print.saws. Added a note to man page for saws.

saws 0.9-6.2 (2019-12-16)

• the geeUOmega function had a call to model.matrix with a "contrasts" argument that was wrong, so that the contrast from the gee output was not necessarily being properly passed down. It is now fixed. Thanks to Muriel Buri for pointing this out, and Allyson Mateja for help with the fix.

saws 0.9-6.1 (2014-01-23)

• had "fake" vignette, so removed .Rnw file. Now just pdf files in the /inst/doc/ directory.

saws 0.9-6.0 (2013-04-05)

• Changed /inst/News.Rd to NEWS
• changed help, no longer says it does Cox regression
• modified mgee function so that if calls the gee function then geeUOmega. The previous method was clunky since it basically copied the code from gee and added some lines. Now I do it a better way by calling the gee function, then using it as an argument in the newly created geeUOmega function which inputs the output from gee. Changed help accordingly.

saws 0.9-5.3 (2012-04-20)

• Changed print.saws file. It had an error when the test matrix was not p X p (where p=number of coefficients).
• Removed dependence on survival. I hope to put adjustment for Cox proportional hazards in the future, but have not done it yet, so there is no dependence.

saws 0.9-3.3 (2010-07-24)

• fix bug when the number of parameters is 1 (see dfCalc.R).