Computes a small set of metrics that flag suspicious matching: high diff density, many fully-different rows, duplicate keys resolved positionally, and similar patterns that often indicate the wrong key was used.
Value
A list with components:
diff_density:n_value_diffs / (n_matched_rows * n_matched_columns)(0-1).n_fully_diff_rows: matched rows where every matched column differs (or, more permissively, >= 80% of columns differ).pct_fully_diff_rows: as proportion of matched rows.n_value_to_na,n_na_to_value: structural NA transitions.dup_keys:TRUEif either side had duplicate key values.dup_positional:TRUEif duplicates were paired positionally (keep_all) — the most error-prone path.row_count_delta:n_comp_rows - n_base_rows(signed).flags: character vector of human-readable warnings (possibly empty).severity:"ok","info","warn", or"critical".
Examples
a <- data.frame(id = c(1, 1, 2), x = c(1, 2, 3))
b <- data.frame(id = c(1, 1, 2), x = c(9, 9, 3))
cmp <- ks_compare(a, b, by = "id", dup_keys = "keep_all")
#> ✖ a vs b — 2 value diffs across 1 column
ks_match_health(cmp)
#> $diff_density
#> [1] 0.3333333
#>
#> $n_fully_diff_rows
#> [1] 0
#>
#> $pct_fully_diff_rows
#> [1] 0
#>
#> $n_value_to_na
#> [1] 0
#>
#> $n_na_to_value
#> [1] 0
#>
#> $dup_keys
#> [1] TRUE
#>
#> $dup_positional
#> [1] TRUE
#>
#> $position_match
#> [1] FALSE
#>
#> $row_count_delta
#> [1] 0
#>
#> $flags
#> [1] "Duplicate keys were paired positionally and diff density is high — the key is likely incomplete. Try adding columns to `by`."
#>
#> $severity
#> [1] "critical"
#>