Changelog • anicheck

anicheck (development version)

The checks handle an aniframe with no rows (#32). check_confidence() aborted with attempt to set an attribute on NULL, and check_na_timing() warned twice about min() and max() having “no non-missing arguments” — neither message mentioning the frame being empty, which is what was actually wrong. An empty frame is a normal product of a grouped pipeline, not a malformed input: dplyr::filter() gives one whenever a group matches nothing.

All three now return an empty check, as check_na_gapsize() already did.
An empty check survives its own summary() and print(). do.call(rbind, list()) is NULL, so a check built from no groups carried a NULL where its methods expect a table, and both failed with invalid argument type. This affected check_na_gapsize() too, whose empty case otherwise worked (#32).

print() on a check object writes its summary to stdout as one block, instead of emitting it as nine messages on stderr. capture.output(print(x)) returned nothing at all before, and suppressMessages() removed the summary entirely — so a pipeline wrapped in suppressMessages() to quiet a repetitive warning also lost its check output. The rendered summary is unchanged (#30).

The minimum anicore is 0.8.0, the first version published under that name — the dependency was renamed without a version constraint, so nothing recorded that a pre-rename aniframe will not do.
The core data structures come from anicore, which is what the aniframe package was renamed to in its 0.8.0 (animovement/anicore#84). The aniframe class keeps its name; only the package providing it changed, so anicore replaces aniframe in Imports and in every aniframe:: call.

Check objects carry the aniframe declarations they were built from, as variables_what and variables_when attributes (animovement/anivis#21). group_cols concatenates identity and temporal context, so a consumer given only that vector cannot tell where one ends and the other begins — “the last grouping column” can land on a session rather than the finest identity. The two fields travel under the names aniframe gives them, so group_cols keeps its meaning and is now explained by the attributes beside it.

First substantial release. anicheck diagnoses movement-data quality, returning objects that anivis knows how to draw.

check_na_timing() summarises where missing values fall, as one row per gap.
check_na_gapsize() tabulates how often each missing-value gap length occurs.
check_confidence() reduces per-keypoint tracking confidence to a compact density grid and five-number summary.
summary() and print() methods for each check, giving a per-group overview.
plot() methods that ensure anivis is installed, prompting if not, and then delegate to it — mirroring the performance / see split in easystats.

Package skeleton. No user-facing functions yet.