Matching Results
Source:R/matching_methods.R, R/matching_plots.R, R/result_methods.R
couplr_result.RdEvery matching function returns an object inheriting from class
"couplr_result":
| Function | Class | Matched units |
match_couples(), cardinality_match() | "matching_result" | pairs |
full_match() | "full_matching_result" | groups |
cem_match() | "cem_result" | strata |
subclass_match() | "subclass_result" | subclasses |
print() gives a short overview and summary() the full report: the
distribution of pair distances, the certificate conditions, the column
generation rounds and the balance constraints of a cardinality matching,
where present. glance() returns the headline numbers as a one-row tibble
and as.data.frame() (or tidy()) the matched units, one row per pair or
per unit. certificate(), potentials() and unmatched_units() read the
certificate, the dual potentials and the unmatched units. Further methods
take any of these results: balance_diagnostics(), join_matched(),
match_data(), as_matchit() and cobalt::bal.tab().
Usage
# S3 method for class 'matching_result'
plot(x, type = c("histogram", "density", "ecdf"), ...)
# S3 method for class 'matching_result'
autoplot(object, type = c("histogram", "density", "ecdf"), ...)
# S3 method for class 'couplr_result'
glance(x, ...)
# S3 method for class 'couplr_result'
tidy(x, ...)
# S3 method for class 'couplr_result'
as.data.frame(x, row.names = NULL, optional = FALSE, ...)
# S3 method for class 'couplr_result'
print(x, ...)
# S3 method for class 'couplr_result'
summary(object, ...)Value
print() returns x invisibly. summary() returns an object of
class "summary.couplr_result". glance() returns a one-row tibble with
columns design, n_left, n_right, n_matched_left,
n_matched_right, n_pairs, n_groups, total_distance, status,
certified and max_suboptimality, plus best_possible, gap,
n_nodes, stopped_on, size_certified (whether the largest size the
constraints admit was proven, which certified also requires of the
balance and distance stages) and size_seconds (how long proving it took,
NA when one network solve answered the call) for a cardinality matching
and n_rounds, candidate_edges and
possible_edges for column generation. as.data.frame() and tidy()
return the matched units.
Examples
set.seed(1)
left <- data.frame(id = 1:5, x = rnorm(5))
right <- data.frame(id = 6:15, x = rnorm(10))
m <- match_couples(left, right, vars = "x", certify = TRUE)
m
glance(m)
as.data.frame(m)
certificate(m)
summary(m)