[WIP] apa_print()-methods for metafor #299
Conversation
Resync to master
| #' } | ||
| #' @family apa_print | ||
| #' | ||
| apa_print.metafor <- function( |
There was a problem hiding this comment.
Unless I'm missing something, for S3-method dispatch to work this functions needs to be named after the S3 class it is ment to operate on. In this case, I think, it should be either the specific rma.uni or more general rma depending on how general this method should be. Right?
|
|
||
| apa_res <- papaja:::apa_print_container() | ||
|
|
||
| apa_res$estimate <- paste0("$", est_name, "$ = ", round(x$b, x$digits[["est"]]), " $95\\% CI$ = $[", round(x$ci.lb, digits = x$digits[["ci"]]), "$, $", round(x$ci.ub, digits = x$digits[["ci"]]), "]$") |
There was a problem hiding this comment.
In terms of a general workflow I try to assemble the table element first and build the rest from it. Do you think this could be a sensible approach here too? It may simplify maintenance and may expose commonalities among functions that could be used further reuse code across methods.
| apa_res <- papaja:::apa_print_container() | ||
|
|
||
| apa_res$estimate <- paste0("$", est_name, "$ = ", round(x$b, x$digits[["est"]]), " $95\\% CI$ = $[", round(x$ci.lb, digits = x$digits[["ci"]]), "$, $", round(x$ci.ub, digits = x$digits[["ci"]]), "]$") | ||
| apa_res$estimate <- setNames(apa_res$estimate, row.names(x$beta)) |
There was a problem hiding this comment.
Nice, I never use setNames() but really should.
There was a problem hiding this comment.
To be honest it's just copied from how the variable names are set in metafor itself.
| apa_res$table$estimate <- broom::tidy(x) | ||
|
|
||
| apa_res$table$heterogeneity <- broom::glance(x) | ||
| names(apa_res$table$heterogeneity) <- c("k", "measure", "method", "I$^2$", "H$^2$", "$\\tau^2$", "$\\tau^2$ se", "Cochran Q$_e$", "Cochran Q$_e$ $p$", "Cochran Q$_m$", "Cochran Q$_m$ $p$") |
There was a problem hiding this comment.
Maybe a nice next step would be to add variable labels using variable_labels().
This is the first attempt at adding a apa_print function for metafor (rma) objects. This will handle both univariate meta-analysis and meta-regression models.
The formatting of the output and documentations needs improvement but this should give you an idea of where we're up to after the hackathon at SIPS2019.