Improve data masking (#138)

* robust data masking * more robust data masking * robust data masking in tests
RMI-PACTA · Sep 18, 2024 · b436300 · b436300
1 parent 847758e
commit b436300
Show file tree

Hide file tree

Showing 22 changed files with 244 additions and 245 deletions.
diff --git a/R/plot_aggregate_loanbooks.R b/R/plot_aggregate_loanbooks.R
@@ -220,7 +220,7 @@ plot_aggregate_loanbooks <- function(config) {
   ### scatter plot alignment by exposure and sector comparison----
   year_scatter_alignment_exposure <- 2027
   region_scatter_alignment_exposure <- region_select
-  currency <- unique(company_aggregated_alignment_net$loan_size_outstanding_currency)
+  currency <- unique(company_aggregated_alignment_net[["loan_size_outstanding_currency"]])
   if (length(by_group) <= 1) {
     if (
       nrow(loanbook_exposure_aggregated_alignment_net) > 0
@@ -533,7 +533,7 @@ plot_aggregate_loanbooks <- function(config) {
   if (length(by_group) == 1) {
     unique_by_group <- company_aggregated_alignment_bo_po %>%
       dplyr::filter(
-        .data$sector == .env$sector_scatter,
+        .data[["sector"]] == .env[["sector_scatter"]],
         !grepl("benchmark_corporate_economy_", !!rlang::sym(by_group))
       ) %>%
       dplyr::pull(!!rlang::sym(by_group)) %>%
@@ -596,7 +596,7 @@ plot_aggregate_loanbooks <- function(config) {
   if (length(by_group) == 1) {
     unique_by_group <- company_aggregated_alignment_bo_po %>%
       dplyr::filter(
-        .data$sector == .env$sector_scatter,
+        .data[["sector"]] == .env[["sector_scatter"]],
         !grepl("benchmark_corporate_economy_", !!rlang::sym(by_group))
       ) %>%
       dplyr::pull(!!rlang::sym(by_group)) %>%
@@ -663,7 +663,7 @@ plot_aggregate_loanbooks <- function(config) {
   if (length(by_group) == 1) {
     unique_by_group <- company_aggregated_alignment_bo_po %>%
       dplyr::filter(
-        .data$sector == .env$sector_scatter,
+        .data[["sector"]] == .env[["sector_scatter"]],
         !grepl("benchmark_corporate_economy_", !!rlang::sym(by_group))
       ) %>%
       dplyr::pull(!!rlang::sym(by_group)) %>%
@@ -726,7 +726,7 @@ plot_aggregate_loanbooks <- function(config) {
   if (length(by_group) == 1) {
     unique_by_group <- company_aggregated_alignment_bo_po %>%
       dplyr::filter(
-        .data$sector == .env$sector_scatter,
+        .data[["sector"]] == .env[["sector_scatter"]],
         !grepl("benchmark_corporate_economy_", !!rlang::sym(by_group))
       ) %>%
       dplyr::pull(!!rlang::sym(by_group)) %>%

diff --git a/R/plot_sankey.R b/R/plot_sankey.R
@@ -48,12 +48,12 @@ plot_sankey <- function(data,
     data_links <- data %>%
       dplyr::mutate(
         group_var = r2dii.plot::to_title(!!rlang::sym(group_var)),
-        middle_node = r2dii.plot::to_title(.data$middle_node)
+        middle_node = r2dii.plot::to_title(.data[["middle_node"]])
       )
     if ("middle_node2" %in% names(data_links)) {
       data_links <- data_links %>%
         dplyr::mutate(
-          middle_node2 = r2dii.plot::to_title(.data$middle_node2)
+          middle_node2 = r2dii.plot::to_title(.data[["middle_node2"]])
         )
     }
   } else {
@@ -62,7 +62,7 @@ plot_sankey <- function(data,
 
   links_1 <- data_links %>%
     dplyr::select(
-      source = .env$group_var,
+      source = .env[["group_var"]],
       target = "middle_node",
       value = "loan_size_outstanding",
       group = "is_aligned"
@@ -71,7 +71,7 @@ plot_sankey <- function(data,
   if ("middle_node2" %in% names(data_links)) {
     links_2 <- data_links %>%
       dplyr::select(
-        .env$group_var,
+        .env[["group_var"]],
         source = "middle_node",
         target = "middle_node2",
         value = "loan_size_outstanding",
@@ -80,7 +80,7 @@ plot_sankey <- function(data,
 
     links_3 <- data_links %>%
       dplyr::select(
-        .env$group_var,
+        .env[["group_var"]],
         source = "middle_node2",
         target = "is_aligned",
         value = "loan_size_outstanding",
@@ -91,7 +91,7 @@ plot_sankey <- function(data,
   } else {
     links_2 <- data_links %>%
       dplyr::select(
-        .env$group_var,
+        .env[["group_var"]],
         source = "middle_node",
         target = "is_aligned",
         value = "loan_size_outstanding",
@@ -102,10 +102,10 @@ plot_sankey <- function(data,
   }
 
   links <- links %>%
-    dplyr::group_by(.data$source, .data$target, .data$group) %>%
-    dplyr::summarise(value = sum(.data$value, na.rm = TRUE)) %>%
+    dplyr::group_by(.data[["source"]], .data[["target"]], .data[["group"]]) %>%
+    dplyr::summarise(value = sum(.data[["value"]], na.rm = TRUE)) %>%
     dplyr::ungroup() %>%
-    dplyr::arrange(.data$source, .data$group) %>%
+    dplyr::arrange(.data[["source"]], .data[["group"]]) %>%
     as.data.frame()
 
   # TODO: colour the companies if fully aligned or not
@@ -114,7 +114,7 @@ plot_sankey <- function(data,
   ) %>%
     dplyr::mutate(
       group = dplyr::case_when(
-        .data$name %in% c("Aligned", "Not aligned", "Unknown") ~ .data$name,
+        .data[["name"]] %in% c("Aligned", "Not aligned", "Unknown") ~ .data[["name"]],
         TRUE ~ "other"
       )
     )

diff --git a/R/plot_scatter.R b/R/plot_scatter.R
@@ -85,18 +85,18 @@ plot_scatter <- function(data,
   if (!is.null(floor_outliers)) {
     data <- data %>%
       dplyr::mutate(
-        buildout = dplyr::if_else(.data$buildout <= .env$floor_outliers, .env$floor_outliers, .data$buildout),
-        phaseout = dplyr::if_else(.data$phaseout <= .env$floor_outliers, .env$floor_outliers, .data$phaseout),
-        net = dplyr::if_else(.data$net <= .env$floor_outliers, .env$floor_outliers, .data$net)
+        buildout = dplyr::if_else(.data[["buildout"]] <= .env[["floor_outliers"]], .env[["floor_outliers"]], .data[["buildout"]]),
+        phaseout = dplyr::if_else(.data[["phaseout"]] <= .env[["floor_outliers"]], .env[["floor_outliers"]], .data[["phaseout"]]),
+        net = dplyr::if_else(.data[["net"]] <= .env[["floor_outliers"]], .env[["floor_outliers"]], .data[["net"]])
       )
     subtitle <- glue::glue("{subtitle}\nThe outliers are displayed on the borders of the plot.", .trim = FALSE)
   }
   if (!is.null(cap_outliers)) {
     data <- data %>%
       dplyr::mutate(
-        buildout = dplyr::if_else(.data$buildout >= .env$cap_outliers, .env$cap_outliers, .data$buildout),
-        phaseout = dplyr::if_else(.data$phaseout >= .env$cap_outliers, .env$cap_outliers, .data$phaseout),
-        net = dplyr::if_else(.data$net >= .env$cap_outliers, .env$cap_outliers, .data$net)
+        buildout = dplyr::if_else(.data[["buildout"]] >= .env[["cap_outliers"]], .env[["cap_outliers"]], .data[["buildout"]]),
+        phaseout = dplyr::if_else(.data[["phaseout"]] >= .env[["cap_outliers"]], .env[["cap_outliers"]], .data[["phaseout"]]),
+        net = dplyr::if_else(.data[["net"]] >= .env[["cap_outliers"]], .env[["cap_outliers"]], .data[["net"]])
       )
     if (is.null(floor_outliers)) {
       subtitle <- glue::glue("{subtitle}\nThe outliers are displayed on the borders of the plot.", .trim = FALSE)
@@ -113,7 +113,7 @@ plot_scatter <- function(data,
     net = c(0, 0, 0)
   )
 
-  p <- ggplot2::ggplot(data, ggplot2::aes(x = .data$buildout, y = .data$phaseout, colour = .data$net)) +
+  p <- ggplot2::ggplot(data, ggplot2::aes(x = .data[["buildout"]], y = .data[["phaseout"]], colour = .data[["net"]])) +
     ggplot2::geom_hline(yintercept = 0, colour = "#c0c0c0") +
     ggplot2::geom_vline(xintercept = 0, colour = "#c0c0c0") +
     ggplot2::geom_line(data = data_net_0) +
@@ -162,7 +162,7 @@ plot_scatter <- function(data,
       size = 3,
       hjust = 0
     ) +
-    ggplot2::geom_point(ggplot2::aes(shape = .data$datapoint)) +
+    ggplot2::geom_point(ggplot2::aes(shape = .data[["datapoint"]])) +
     ggplot2::scale_x_continuous(
       name = "Deviation from scenario value\nfor low-carbon technologies build-out",
       labels = scales::percent,

diff --git a/R/plot_scatter_alignment_exposure.R b/R/plot_scatter_alignment_exposure.R
@@ -40,9 +40,9 @@ plot_scatter_alignment_exposure <- function(data,
     data <- data %>%
       dplyr::mutate(
         exposure_weighted_net_alignment = dplyr::if_else(
-          .data$exposure_weighted_net_alignment <= .env$floor_outliers,
-          .env$floor_outliers,
-          .data$exposure_weighted_net_alignment
+          .data[["exposure_weighted_net_alignment"]] <= .env[["floor_outliers"]],
+          .env[["floor_outliers"]],
+          .data[["exposure_weighted_net_alignment"]]
         )
       )
   }
@@ -51,9 +51,9 @@ plot_scatter_alignment_exposure <- function(data,
     data <- data %>%
       dplyr::mutate(
         exposure_weighted_net_alignment = dplyr::if_else(
-          .data$exposure_weighted_net_alignment >= .env$cap_outliers,
-          .env$cap_outliers,
-          .data$exposure_weighted_net_alignment
+          .data[["exposure_weighted_net_alignment"]] >= .env[["cap_outliers"]],
+          .env[["cap_outliers"]],
+          .data[["exposure_weighted_net_alignment"]]
         )
       )
   }
@@ -68,11 +68,11 @@ plot_scatter_alignment_exposure <- function(data,
   }
 
   plot <- data %>%
-    dplyr::mutate(sector = tools::toTitleCase(.data$sector)) %>%
+    dplyr::mutate(sector = tools::toTitleCase(.data[["sector"]])) %>%
     ggplot2::ggplot(
       ggplot2::aes(
-        x = .data$sum_loan_size_outstanding,
-        y = .data$exposure_weighted_net_alignment,
+        x = .data[["sum_loan_size_outstanding"]],
+        y = .data[["exposure_weighted_net_alignment"]],
         color = !!rlang::sym(group_var)
       )
     ) +

diff --git a/R/plot_scatter_animated.R b/R/plot_scatter_animated.R
@@ -81,25 +81,25 @@ plot_scatter_animated <- function(data,
   if (!is.null(floor_outliers)) {
     data <- data %>%
       dplyr::mutate(
-        buildout = dplyr::if_else(.data$buildout <= .env$floor_outliers, .env$floor_outliers, .data$buildout),
-        phaseout = dplyr::if_else(.data$phaseout <= .env$floor_outliers, .env$floor_outliers, .data$phaseout),
+        buildout = dplyr::if_else(.data[["buildout"]] <= .env[["floor_outliers"]], .env[["floor_outliers"]], .data[["buildout"]]),
+        phaseout = dplyr::if_else(.data[["phaseout"]] <= .env[["floor_outliers"]], .env[["floor_outliers"]], .data[["phaseout"]]),
         net = dplyr::if_else(
-          .data$buildout <= .env$floor_outliers | .data$phaseout <= .env$floor_outliers,
-          .data$buildout + .data$phaseout,
-          .data$net
+          .data[["buildout"]] <= .env[["floor_outliers"]] | .data[["phaseout"]] <= .env[["floor_outliers"]],
+          .data[["buildout"]] + .data[["phaseout"]],
+          .data[["net"]]
         ) # net is a sum of buildout and phaseout
       )
     subtitle <- glue::glue("{subtitle}\nThe outliers are displayed on the borders of the plot.", .trim = FALSE)
   }
   if (!is.null(cap_outliers)) {
     data <- data %>%
       dplyr::mutate(
-        buildout = dplyr::if_else(.data$buildout >= .env$cap_outliers, .env$cap_outliers, .data$buildout),
-        phaseout = dplyr::if_else(.data$phaseout >= .env$cap_outliers, .env$cap_outliers, .data$phaseout),
+        buildout = dplyr::if_else(.data[["buildout"]] >= .env[["cap_outliers"]], .env[["cap_outliers"]], .data[["buildout"]]),
+        phaseout = dplyr::if_else(.data[["phaseout"]] >= .env[["cap_outliers"]], .env[["cap_outliers"]], .data[["phaseout"]]),
         net = dplyr::if_else(
-          .data$buildout >= .env$cap_outliers | .data$phaseout >= .env$cap_outliers,
-          .data$buildout + .data$phaseout,
-          .data$net
+          .data[["buildout"]] >= .env[["cap_outliers"]] | .data[["phaseout"]] >= .env[["cap_outliers"]],
+          .data[["buildout"]] + .data[["phaseout"]],
+          .data[["net"]]
         ) # net is a sum of buildout and phaseout
       )
     if (is.null(floor_outliers)) {

diff --git a/R/plots.R b/R/plots.R
@@ -24,9 +24,9 @@ plot_match_success_rate <- function(data,
   stop_if_not_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)
+    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(
@@ -139,30 +139,30 @@ generate_individual_outputs <- function(data,
 
   data <- data %>%
     dplyr::filter(
-      .data[[by_group]] == .env$by_group_value,
-      .data[["scenario_source"]] == .env$scenario_source,
-      .data[["region"]] == .env$region,
-      .data[["sector"]] %in% .env$sector
+      .data[[by_group]] == .env[["by_group_value"]],
+      .data[["scenario_source"]] == .env[["scenario_source"]],
+      .data[["region"]] == .env[["region"]],
+      .data[["sector"]] %in% .env[["sector"]]
     )
 
   matched_prioritized <- matched_prioritized %>%
     dplyr::filter(
-      .data[[by_group]] == .env$by_group_value,
-      .data[["sector"]] %in% .env$sector
+      .data[[by_group]] == .env[["by_group_value"]],
+      .data[["sector"]] %in% .env[["sector"]]
     )
 
   if (target_type == "tms") {
     # plot tech mix for given sector
     data_techmix <- data %>%
       dplyr::filter(
-        .data[["metric"]] %in% c("projected", "corporate_economy", .env$target_scenario),
-        dplyr::between(.data[["year"]], .env$start_year, .env$start_year + .env$time_horizon)
+        .data[["metric"]] %in% c("projected", "corporate_economy", .env[["target_scenario"]]),
+        dplyr::between(.data[["year"]], .env[["start_year"]], .env[["start_year"]] + .env[["time_horizon"]])
       ) %>%
       dplyr::mutate(
         label = dplyr::case_when(
           .data[["metric"]] == "projected" ~ "Portfolio",
           .data[["metric"]] == "corporate_economy" ~ "Corporate Economy",
-          .data[["metric"]] == .env$target_scenario ~ glue::glue("{r2dii.plot::to_title(toupper(.env$scenario))} Scenario")
+          .data[["metric"]] == .env[["target_scenario"]] ~ glue::glue("{r2dii.plot::to_title(toupper(.env$scenario))} Scenario")
         )
       ) %>%
       r2dii.plot::prep_techmix(
@@ -207,13 +207,13 @@ generate_individual_outputs <- function(data,
 
     # plot trajectory charts for all available techs in given sector
     technologies_in_sector <- r2dii.data::increasing_or_decreasing %>%
-      dplyr::filter(.data[["sector"]] == .env$sector) %>%
+      dplyr::filter(.data[["sector"]] == .env[["sector"]]) %>%
       dplyr::pull(.data[["technology"]])
 
     technologies_to_plot <- data %>%
       dplyr::filter(
-        .data[["metric"]] == .env$target_scenario,
-        .data[["technology"]] %in% .env$technologies_in_sector
+        .data[["metric"]] == .env[["target_scenario"]],
+        .data[["technology"]] %in% .env[["technologies_in_sector"]]
       ) %>%
       dplyr::distinct(.data[["technology"]]) %>%
       dplyr::arrange(.data[["technology"]]) %>%
@@ -222,8 +222,8 @@ generate_individual_outputs <- function(data,
     for (i in 1:length(technologies_to_plot)) {
       data_trajectory <- data %>%
         dplyr::filter(
-          .data[["technology"]] == .env$technologies_to_plot[i],
-          dplyr::between(.data[["year"]], .env$start_year, .env$start_year + .env$time_horizon)
+          .data[["technology"]] == .env[["technologies_to_plot"]][i],
+          dplyr::between(.data[["year"]], .env[["start_year"]], .env[["start_year"]] + .env[["time_horizon"]])
         ) %>%
         r2dii.plot::prep_trajectory(
           convert_label = r2dii.plot::recode_metric_trajectory,
@@ -271,15 +271,15 @@ generate_individual_outputs <- function(data,
       dplyr::filter(
         dplyr::between(
           .data[["year"]],
-          .env$start_year,
-          .env$start_year + .env$time_horizon)
+          .env[["start_year"]],
+          .env[["start_year"]] + .env[["time_horizon"]])
       ) %>%
       dplyr::filter(
         .data[["emission_factor_metric"]] %in% c(
           "projected",
           "corporate_economy",
-          .env$target_scenario,
-          .env$adjusted_scenario
+          .env[["target_scenario"]],
+          .env[["adjusted_scenario"]]
         )
       ) %>%
       dplyr::mutate(
@@ -288,8 +288,8 @@ generate_individual_outputs <- function(data,
           levels = c(
             "projected",
             "corporate_economy",
-            .env$target_scenario,
-            .env$adjusted_scenario
+            .env[["target_scenario"]],
+            .env[["adjusted_scenario"]]
           )
         )
       ) %>%
@@ -350,7 +350,7 @@ generate_individual_outputs <- function(data,
     dplyr::select(
       dplyr::all_of(
         c(
-          .env$by_group, "name_abcd", "sector_abcd", "loan_size_outstanding",
+          .env[["by_group"]], "name_abcd", "sector_abcd", "loan_size_outstanding",
           "loan_size_outstanding_currency", "loan_size_credit_limit",
           "loan_size_credit_limit_currency"
         )