Changelog • datascan

datascan (development version)

cols_missing() and rows_missing() with cutoff = 0 now return only the columns or rows that have at least one missing value, instead of all of them.
concurrence_matrix() and concurrence_table() now treat NA as a level when na.rm = FALSE, instead of silently dropping it. As a result, is_nested(), is_complete() and cols_nested() with na.rm = FALSE now take missing values into account.
concurrence_table() now shows the name of the grouping variable in its header, instead of always saying “environments”.
cols_bijective() no longer reports constant columns as bijective with each other.
tidyr is no longer a dependency.

CRAN release: 2026-10-06