The hardware and bandwidth for this mirror is donated by dogado GmbH, the Webhosting and Full Service-Cloud Provider. Check out our Wordpress Tutorial.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]dogado.de.
MetaLab data now lives in versioned, citable releases on Redivis (https://stanford.redivis.com/datasets/81tq-8dp5ge6b9), and metalabr reads them by default. Every data-access call announces which release it used.
get_metalab_data() now reads released data from Redivis
by default, with a version argument ("current"
by default, or a release name like "2023.1" to pin). The
legacy live-Google-Sheets path is still available by passing dataset
metadata (get_metalab_data(metadata)); it is intended for
curators building releases.get_metalab_metadata() likewise reads the released
dataset registry by default (with
sheet_status/data_as_of provenance columns);
passing a datasets.yaml path/URL retains the legacy
behavior.get_current_metalab_data() now downloads the current
release served by the MetaLab site instead of a snapshot committed to
GitHub in 2023, and returns the effect-size data frame — like every
other data-access function — with the release announced in a message and
recorded in the metalab_release attribute. (metalabr 0.x
instead loaded objects into the global environment, which CRAN policy
disallows; use get_metalab_metadata() for what was the
dataset_info object.)get_metalab_data(imputation_seed = 111), the historical
default, reproducing released data exactly), and the session’s RNG state
is saved and restored around the imputation.get_metalab_versions() lists releases and their
Redivis versions.vignette("metalabr"))
covering released-data access and version pinning.NULL) rather than an error.metalab_scatter_plot() was unusable outside the retired
Shiny app (it referenced reactive objects); it now has a working
data-in/ggplot-out interface consistent with the other plot
functions.metalab_violin_plot() supports every effect-size type
(previously it errored for anything but Hedges’ g) and handles multiple
moderators.metalab_forest_plot() fits its model once instead of
four times, and its moderator handling works outside the app.metalab_funnel_plot() works with moderators outside the
app (it previously referenced free variables that only existed in the
app environment).metalab_funnel_test(): an Egger-style asymmetry
test adapted to the multilevel model (√vᵢ as a fixed-effect moderator).
The legacy app’s regtest() call always errored (invisibly)
on multilevel models.langcog
package.viz_app() is deprecated: the interactive applications
are superseded by the in-browser visualizations at https://metalab.stanford.edu. The internal Shiny app
code (visualization, power analysis, validation) and the abandoned
Google-Drive revision machinery were removed.library(metalabr) works without attaching dplyr first
(fixes #6).dplyr::add_rownames() (defunct) and other retired
tidyverse idioms were replaced; the pipeline runs on current tidyverse
releases.License field replaced with MIT + file
LICENSE.rule.type is named rule_type (Redivis
does not allow dots in column names). Data frames loaded via
get_current_metalab_data() keep the legacy name.These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.
Health stats visible at Monitor.