make tidyverse syntax compliant with tidyverse guidelines for packages

.Rbuildignore

@@ -1,6 +1,12 @@
 ^renv$
 ^renv\.lock$
 ^LICENSE\.md$
+^DISCLAIMER\.md$
+^CONTRIBUTING\.md$
+^open_practices\.md$
+^rules_of_behavior\.md$
+^code-of-conduct\.md$
+^thanks\.md$
 ^_pkgdown\.yml$
 ^docs$
 ^pkgdown$

NAMESPACE

@@ -31,4 +31,5 @@ export(update_hub)
 export(us_loc_abbr_to_code)
 export(us_loc_code_to_abbr)
 export(validate_base_forecasts)
+importFrom(rlang,":=")
 importFrom(rlang,.data)

R/create_table_for_scoring.R

@@ -57,11 +57,13 @@ update_hub <- function(hub_path) {
 gather_hub_forecast_data <- function(hub_path) {
   hub_connection <- hubData::connect_hub(hub_path)
   forecasts <- hub_connection |>
-    dplyr::filter(output_type == "quantile") |>
-    dplyr::filter(horizon >= 0) |>
-    dplyr::rename(prediction = value) |>
-    dplyr::rename(forecast_date = reference_date) |>
-    dplyr::rename(quantile = output_type_id) |>
+    dplyr::filter(.data$output_type == "quantile") |>
+    dplyr::filter(.data$horizon >= 0) |>
+    dplyr::rename(
+      "prediction" = "value",
+      "forecast_date" = "reference_date",
+      "quantile" = "output_type_id"
+    ) |>
     dplyr::collect()
   return(forecasts)
 }
@@ -115,7 +117,7 @@ gather_target_data <- function(hub_path,
   if (truth_data_path |> fs::file_exists()) {
     target_data <-
       readr::read_csv(truth_data_path) |>
-      dplyr::rename(true_value = value)
+      dplyr::rename("true_value" = "value")
     return(target_data)
   } else {
     cfl <- truth_data_path
@@ -146,23 +148,23 @@ create_table_for_scoring <- function(hub_path) {
     forecasts_only,
     target_data,
     by = dplyr::join_by(
-      location == location,
-      target_end_date == date
+      "location" == "location",
+      "target_end_date" == "date"
     )
   ) |>
-    dplyr::rename(model = model_id) |>
-    dplyr::mutate(target_end_date = as.Date(target_end_date)) |>
-    dplyr::mutate(quantile = as.numeric(quantile)) |>
+    dplyr::rename("model" = "model_id") |>
+    dplyr::mutate("target_end_date" = as.Date(.data$target_end_date)) |>
+    dplyr::mutate("quantile" = as.numeric(.data$quantile)) |>
     dplyr::select(
-      forecast_date,
-      target_end_date,
-      horizon,
-      location,
-      quantile,
-      prediction,
-      location_name,
-      true_value,
-      model
+      "forecast_date",
+      "target_end_date",
+      "horizon",
+      "location",
+      "quantile",
+      "prediction",
+      "location_name",
+      "true_value",
+      "model"
     )
 
   return(forecasts_and_targets)

R/daily_to_epiweekly.R

@@ -66,17 +66,17 @@ daily_to_epiweekly <- function(tidy_daily_trajectories,
 
   grouped_df <- tidy_daily_trajectories |>
     dplyr::mutate(
-      epiweek = lubridate::epiweek(.data[[!!date_col]]),
-      epiyear = lubridate::epiyear(.data[[!!date_col]])
+      "epiweek" = lubridate::epiweek(.data[[!!date_col]]),
+      "epiyear" = lubridate::epiyear(.data[[!!date_col]])
     ) |>
     dplyr::group_by(
-      epiweek,
-      epiyear,
+      .data$epiweek,
+      .data$epiyear,
       dplyr::across(tidyselect::all_of(!!id_cols))
     )
 
   n_elements <- grouped_df |>
-    dplyr::summarise(n_elements = dplyr::n()) |>
+    dplyr::summarise("n_elements" = dplyr::n()) |>
     dplyr::pull()
 
   ## check that no weeks have more than

R/filter_for_scoring.R

@@ -39,7 +39,7 @@ filter_for_scoring <- function(forecasts_and_targets,
       "Filtering out forecasts without corresponding truth data..."
     )
     to_score <- to_score |>
-      dplyr::filter(!is.na(true_value))
+      dplyr::filter(!is.na(.data$true_value))
   }
 
   # filter forecasts to score
@@ -48,21 +48,21 @@ filter_for_scoring <- function(forecasts_and_targets,
       "Filtering out forecast dates before {min_forecast_date}"
     )
     to_score <- to_score |>
-      dplyr::filter(forecast_date >= as.Date(!!min_forecast_date))
+      dplyr::filter(.data$forecast_date >= as.Date(!!min_forecast_date))
   }
   if (!is.null(max_forecast_date)) {
     cli::cli_inform(
       "Filtering out forecast dates after {max_forecast_date}"
     )
     to_score <- to_score |>
-      dplyr::filter(forecast_date <= as.Date(!!max_forecast_date))
+      dplyr::filter(.data$forecast_date <= as.Date(!!max_forecast_date))
   }
   if (!is.null(horizons)) {
     cli::cli_inform(
       "Filtering out horizons not in {horizons}"
     )
     to_score <- to_score |>
-      dplyr::filter(horizon %in% c(!!horizons))
+      dplyr::filter(.data$horizon %in% c(!!horizons))
   }
   if (!is.null(locations)) {
     cli::cli_inform(
@@ -73,14 +73,14 @@ filter_for_scoring <- function(forecasts_and_targets,
       locations
     )
     to_score <- to_score |>
-      dplyr::filter(location %in% c(!!loc_codes))
+      dplyr::filter(.data$location %in% c(!!loc_codes))
   }
   if (!is.null(models)) {
     cli::cli_inform(
       "Filtering out models not in {models}"
     )
     to_score <- to_score |>
-      dplyr::filter(model %in% c(!!models))
+      dplyr::filter(.data$model %in% c(!!models))
   }
 
   return(to_score)

R/forecasttools-package.R

@@ -2,6 +2,7 @@
 "_PACKAGE"
 
 ## usethis namespace: start
+#' @importFrom rlang :=
 #' @importFrom rlang .data
 ## usethis namespace: end
 NULL

R/simple_bottom_up.R

@@ -94,10 +94,10 @@ count_trajectories <- function(base_forecasts,
                                location_col,
                                date_col) {
   number_of_sampled_trajs <- base_forecasts |>
-    dplyr::group_by(.data[[location_col]], .data[[date_col]]) |>
-    dplyr::summarise(n_sample_trajs = dplyr::n()) |>
+    dplyr::group_by(.data[[!!location_col]], .data[[!!date_col]]) |>
+    dplyr::summarise("n_sample_trajs" = dplyr::n()) |>
     dplyr::ungroup() |>
-    dplyr::select(dplyr::all_of(location_col), n_sample_trajs) |>
+    dplyr::select(dplyr::all_of(!!location_col), "n_sample_trajs") |>
     dplyr::distinct()
   return(number_of_sampled_trajs)
 }