The scalar summaries of missingness: how much of the table is missing,
how many rows are complete, and how many columns have any or only missing
values. Use it as a first look at a new extract, or to report data
completeness in a methods section;
misspat() then shows which
columns are missing together.
Value
A named numeric vector: nrows, n_cols,
n_missing (missing cells), pct_missing (percent of cells),
n_complete_rows, pct_complete_rows (percent of rows),
n_cols_any_missing (columns with at least one missing value) and
n_cols_all_missing (columns with no value at all).
Details
A cell is missing when is.na() is TRUE (so NaN
counts, and an empty string does not). A row is complete when no cell in it
is missing (stats::complete.cases()). Percentages are of all cells
and of all rows. A table with no rows gives zeros. Columns are taken by
position, so duplicate or empty column names are fine. Errors when
data is not a data frame or has no columns.
See also
misspat() for which
columns are missing together,
profcols() for per-column rates.
Examples
df <- data.frame(a = c(1, NA, 3), b = c(NA, NA, 3), c = 1:3)
misssmry(df)
#> n_rows n_cols n_missing pct_missing
#> 3.00000 3.00000 3.00000 33.33333
#> n_complete_rows pct_complete_rows n_cols_any_missing n_cols_all_missing
#> 1.00000 33.33333 2.00000 0.00000
# A complete table is all zeros but for its dimensions.
misssmry(data.frame(x = 1:3, y = 4:6))
#> n_rows n_cols n_missing pct_missing
#> 3 2 0 0
#> n_complete_rows pct_complete_rows n_cols_any_missing n_cols_all_missing
#> 3 100 0 0
# a wholly empty column
ms <- misssmry(data.frame(x = 1:4, note = NA))
ms[c("pct_missing", "n_cols_all_missing")]
#> pct_missing n_cols_all_missing
#> 50 1
# a vector is not a table
try(misssmry(c(1, NA, 3)))
#> Error : misssmry: `data` must be a data frame