Inspect Atlas Coverage and Observation Counts
summary.offline_atlas.RdSummarize supplied areas and relationships before exporting. Return inspectable tables without changing the atlas or inferring absent rows.
Details
Area counts are grouped by original variable and period. Static records have NA in the time column. Periods sort numerically when all can be read as finite numbers, and otherwise lexically. A multi-variable panel counts each area-period once in area_observations; each variable has its own diagnostic table rows.
Missing is the total number of NA values. Unmatched counts areas flagged matched = FALSE. Unavailable counts NA values in matched areas. Thus missing equals unmatched plus unavailable; these columns are not three disjoint groups. Observed counts nonmissing values. Negative, zero and positive counts apply only to numeric variables and weighted relationships. They are NA for categories and unweighted ties, whose implicit unit weights are not measurements.
Category counts retain original category names and all declared levels, including unused levels with zero observations. Missing categories appear in the areas table, not as a fabricated category. Relationships are grouped by supplied period and layer. Their endpoints were validated when added, so area matching diagnostics do not apply. Omitted observed pairs are not counted as zero or missing. A summary cannot report source rows that were discarded before constructing the atlas, or identify which periods ought to have been supplied.
The summary uses all supplied periods and data, regardless of initial browser filters or legend aliases. It performs no download or geometry conversion.
Value
A summary_offline_atlas list containing title, coordinates, entities (unique node and area counts and relationship row count), periods, weighted, directed, coverage, area_observations, areas, categories, relationships, view, and preset. The three diagnostic tables are data frames, including when empty. The print method returns its input invisibly.
Examples
shapes <- example_data("world_paths")[1:3, ]
shapes$value <- c(-2, 0, NA_real_)
shapes$matched <- c(TRUE, TRUE, FALSE)
x <- atlas("Coverage example") |>
add_areas(shapes, values = "value", matched = "matched",
breaks = list(c(0, 10)))
audit <- summary(x)
audit$areas
#> variable label type time total observed missing unmatched unavailable
#> 1 value value numeric <NA> 3 2 1 1 0
#> negative zero positive
#> 1 1 1 0
print(audit, max_rows = 3)
#> Coverage example
#> 0 nodes, 0 relationship rows, 3 areas
#> Periods: none (static)
#>
#> Area observations by variable and period
#> variable label type time total observed missing unmatched unavailable
#> value value numeric <NA> 3 2 1 1 0
#> negative zero positive
#> 1 1 0
#>
#> Missing includes unmatched areas and matched-but-unavailable observations.
#> Numeric sign counts are NA for categories and unweighted ties. No absent rows are inferred.