Consider exporting plot_match_success_rate from r2dii.plot (or r2dii.match?)

[jdhoffa]

This function plots the matching success rate, which is useful to all users that want to run matching (not just multi.loanbook).

We might consider exporting it from a different repo so that it can be easily accessed by these users.

pacta.multi.loanbook/R/plots.R

Lines 1 to 94 in cab8d28

	plot_match_success_rate <- function(data,
	metric_type = c("absolute", "relative"),
	match_success_type = c("n", "outstanding", "credit_limit"),
	currency,
	by_group) {
	# validate inputs
	assert_inherits(data, "data.frame")
	expected_cols <- c(
	by_group,
	"sector",
	"matched",
	"match_success_type",
	"match_success_rate",
	"metric_type"
	)
	assert_expected_columns(data, expected_cols, desc = "Input")
	assert_inherits(metric_type, "character")
	assert_inherits(match_success_type, "character")
	assert_length(currency, 1L)
	assert_inherits(currency, "character")
	data <- data %>%
	dplyr::filter(.data[["sector"]] != "not in scope") %>%
	dplyr::filter(.data[["metric_type"]] == .env[["metric_type"]]) %>%
	dplyr::filter(.data[["match_success_type"]] == .env[["match_success_type"]])
	# plot design
	fill_scale <- c(
	"Matched" = "#00c082",
	"Not matched" = "#a63d57"
	)
	theme_match_success <- ggplot2::theme(
	legend.position = "top",
	legend.title = ggplot2::element_blank(),
	axis.text.x = ggplot2::element_text(angle = 90, vjust = 0.5, hjust=1)
	)
	# plot description
	title <- r2dii.plot::to_title(glue::glue("{metric_type} Match Success Rate"))
	if (match_success_type == "n") {
	subtitle <- "number of loans by sector"
	} else if (match_success_type == "outstanding") {
	subtitle <- "loan size outstanding by sector"
	} else {
	subtitle <- "credit limit by sector"
	}
	subtitle <- r2dii.plot::to_title(glue::glue("{subtitle} and {by_group}"))
	if (match_success_type == "n") {
	y_label <- "Match success rate (n)"
	} else if (match_success_type == "outstanding") {
	y_label <- "Match success rate: loan size outstanding"
	} else {
	y_label <- "Match success rate: credit limit"
	}
	if (metric_type == "absolute") {
	y_label <- glue::glue("{y_label} (in {currency})")
	} else {
	y_label <- glue::glue("{y_label} (share of total)")
	}
	# plot
	plot <- data %>%
	ggplot2::ggplot(
	ggplot2::aes(
	x = r2dii.plot::to_title(.data[["sector"]]),
	y = .data[["match_success_rate"]],
	fill = .data[["matched"]]
	)
	) +
	ggplot2::geom_col(
	position = ggplot2::position_stack(reverse = TRUE)
	) +
	ggplot2::scale_fill_manual(values = fill_scale) +
	ggplot2::labs(
	x = "Sector",
	y = y_label,
	title = title,
	subtitle = subtitle
	) +
	ggplot2::theme_bw() +
	theme_match_success +
	ggplot2::facet_wrap(ggplot2::vars(r2dii.plot::to_title(.data[[by_group]])))
	plot
	}

[jdhoffa]

Since this package already depends on r2dii.plot and r2dii.match we could still just call it from here.