Reconcile two sources under an explicit matching schema

The reconciliation step of the MRM: joins a primary and a secondary source on identifier keys, optionally tolerating bounded disagreement on numeric fields (e.g. dates recorded a day apart), and reports the match rate, the orphans on each side, and field-level conflicts among matched records.

Usage

morie_mrm_reconcile(
  primary,
  secondary,
  keys,
  compare = NULL,
  numeric_tolerance = 0
)

Arguments

primary, secondary

Data frames (or morie_mrm_dataset objects).

keys

Character vector of join-key column names present in both sources.

compare

Optional character vector of shared columns to check for field-level conflicts among matched rows.

numeric_tolerance

Absolute tolerance under which numeric disagreements in compare columns are NOT conflicts (default 0).

Value

An object of class morie_mrm_reconciliation: list with matched (merged data frame), unmatched_primary, unmatched_secondary, conflicts (long data frame: key, field, primary_value, secondary_value), match_rate, schema.

Examples

a <- data.frame(id = 1:5, y = c(1, 2, 3, 4, 5))
b <- data.frame(id = c(1:4, 9), y = c(1, 2, 3.5, 4, 9))
r <- morie_mrm_reconcile(a, b, keys = "id", compare = "y")
r$match_rate
#> [1] 0.8
nrow(r$conflicts)
#> [1] 1