LoginSign Up
← Back to all sparks
R

rempsyc

ANALYTICS
Velocity0.0

R package rempsyc by rempsyc — release notes from GitHub.

github.com/rempsyc/rempsyc ↗

Publication-ready psychology tables and plots, tracking APA style as closely as the software allows.

apa-formattingpsychology-researchstatistical-tablesggplot2easystatsr-package
◆Current state
rempsyc produces APA-formatted tables and figures for psychology research — nice_table() for results tables, plus plotting helpers for scatter plots, violin plots, densities and simple slopes. Its releases are CRAN submissions that bundle a long run of development versions, so each entry reads as a digest rather than a single change. The most recent, 0.2.0, added point labelling and per-group correlation statistics to nice_scatter and fixed nice_lm() failing on factor covariates with more than two levels.
◆Where it's heading
Two forces drive this package and neither is its own roadmap. The first is APA style: when the 7th edition advised against beta for standardized coefficients, the package switched its output to italic b with an asterisk. The second is the surrounding ecosystem — formatting is aligned to what lavaanExtra and afex produce, contrast handling was delegated to easystats' modelbased, and Excel correlation matrix export was handed entirely to the correlation package to cut maintenance.
◆Prediction
The pattern of delegating functionality to specialist packages while keeping the formatting layer is well established and likely continues. Because releases bundle many small dev versions, the next one will probably again mix plotting refinements with fixes surfaced by upstream changes.

◆Recent moves

  1. 11mo ago

    Point labels and per-group correlations added to nice_scatter

    The release notes open with continuous integration housekeeping, but the substance sits at the end: nice_scatter gains has.ids for labelling individual points via ggrepel and has.group.r and has.group.p for reporting correlations per group. A nice_lm() fix for factor covariates with more than two levels resolves a failure that would have blocked common designs outright.

    View source ↗
  2. 1y ago

    Excel correlation export delegated to the correlation package

    cormatrix_excel becomes a thin call to correlation::cormatrix_to_excel, explicitly to reduce maintenance. A clear instance of the package's habit of keeping the formatting layer and handing the computation to whoever specialises in it.

    View source ↗
  3. 2y ago

    Table spacing control and a fix for name collision with afex

    A digest release whose recurring theme is friction with neighbouring packages: the nice_table class collided with afex's print method, paired t-tests needed a workaround, and plot_means_over_time was extended past five groups. The new spacing argument lets tables switch between double and single spacing as journals require.

    View source ↗
  4. 2y ago

    Standardized coefficients switch to APA 7th edition b* notation

    The clearest example of the package following the style manual rather than its users' habits: APA 7 advises against beta for standardized coefficients, so output changed to italic b with an asterisk. Z statistics also moved to two digits to match Wald Z reporting in structural equation modelling output.

    View source ↗
  5. 2y ago

    Legend and standardization-check fixes

    A small CRAN submission covering a redundant legend in nice_varplot and a bug when checking standardization on data containing character variables. No new capability.

    View source ↗
  6. 2y ago

    nice_table starts coercing model objects automatically

    Automatically converting objects such as emmGrid to data frames, and renaming t.ratio to t, removes a manual step between fitting a model and formatting its output. Small individually, but this accumulation of input coercions is what makes the table function usable across the many model classes psychology workflows produce.

    View source ↗