Forum: developers
Monitor Forum | 
Start New Thread
| Things to do for 1.0 [ Reply ] By: Richard Morey on 2013-03-08 19:09 | [forum:39135] |
|
1. change default scale for one sample t test so that it matches ANOVA (the scale was changed for the two-sample, but did not need changed for one-sample) COMMITTED 2. Make sure knitr vignettes work for R 3.0 3. change quadform use to quadform2 in c code, for speed COMMITTED 4. decide what to do about subject interactions - new function, or arguments to anovaBF? 5. Make glm possible (allow mixing of continuous and categorical covariates) COMMITTED, resulting Bayes factors check out. Could use more testing. 6. simple interface, with shiny? 7. Clean up output format (print method for bfObject) 8. Make labels for g parameters in Gibbs sampler output clear about which parameter they're for; g/(1+g) is useful as a sort of "partial eta-squared"-like measure |
|
