Since the beginning of this year, I've made 84 commits to the 'development' version of the metafor package. Some recent additions include adding a
vif() function to obtain 'variance inflation factors' to diagnose potential collinearity among the predictor variables in meta-regression models, being able to specify vectors for the
bg arguments in
funnel(), some additional measures for
escalc(), and various improvements to the error messages and documentation. I really should get around to releasing a new version via CRAN soon. A good New Year's resolution!
Despite the lack of updates here, I've been working (off and on) on the metafor package over the past six months and lots of updates have now started to accumulate once again in the 'development' version of the package (which you can always install as described here). Some of the updates include various improvements to the influence and outlier diagnostics for models fitted with
rma.mv(), the addition of continuous-time autoregressive random effects and spatial correlation structures for
rma.mv() models, the possibility to model correlated random effects for any number of predictors, including continuous ones (i.e., this allows for 'random slopes'), and the ability to show styled/colored output when the crayon package is loaded (but note that this only works on a 'proper' terminal, but also in RStudio). It's probably time to push another release to CRAN soon.