.drop fixes

R/end_rains.R

@@ -8,6 +8,7 @@
 #' @param rain \code{character(1)} The name of the rainfall column in \code{data} to apply the function to.
 #' @param doy \code{character(1)} The name of the day of year column in \code{data} to apply the function to. If \code{NULL} it will be created using the \code{date_time} variable.
 #' @param s_start_doy \code{numerical(1)} Default `NULL` (if `NULL`, `s_start_doy = 1`. The day of year to state is the first day of year.
+#' @param drop \code{logical(1)} default `TRUE`. Whether to drop years where there are `NA` data for the rainfall.
 #' @param start_day \code{numerical(1)} The first day to calculate from in the year (1-366).
 #' @param end_day \code{numerical(1)} The last day to calculate to in the year (1-366).
 #' @param output \code{character(1)} Whether to give the start of rains by day of year (doy), date, or both. Default `"doy"`.
@@ -20,7 +21,7 @@
 #' @examples #TODO#
 #' # is same as R-Instat
 end_rains <- function(data, date_time, station = NULL, year = NULL, rain = NULL,
-                      doy = NULL,  s_start_doy = NULL,
+                      doy = NULL,  s_start_doy = NULL, drop = TRUE,
                       start_day = 1, end_day = 366, output = c("doy", "date", "both"),
                       interval_length = 1, min_rainfall = 10){
 
@@ -66,15 +67,15 @@ end_rains <- function(data, date_time, station = NULL, year = NULL, rain = NULL,
   data[[year]] <- factor(data[[year]])
   if (!is.null(station)){
     end_of_rains <- data %>% 
-      dplyr::group_by(.data[[station]], .drop = FALSE) 
+      dplyr::group_by(.data[[station]], .drop = drop) 
   } else {
     end_of_rains <- data
   }
 
   end_of_rains <- end_of_rains %>%
     dplyr::mutate(roll_sum_rain = RcppRoll::roll_sumr(x = .data[[rain]], n = interval_length, fill = NA, na.rm = FALSE)) %>%
     dplyr::filter((roll_sum_rain > min_rainfall) | is.na(x=roll_sum_rain)) %>% 
-    dplyr::group_by(.data[[year]], .add = TRUE, .drop = FALSE) %>%
+    dplyr::group_by(.data[[year]], .add = TRUE, .drop = drop) %>%
     dplyr::filter(.data[[doy]] >= start_day & .data[[doy]] <= end_day, .preserve = TRUE)
 
   if (output == "doy"){

R/end_season.R

@@ -8,6 +8,7 @@
 #' @param rain \code{character(1)} The name of the rainfall column in \code{data} to apply the function to.
 #' @param doy \code{character(1)} The name of the day of year column in \code{data} to apply the function to. If \code{NULL} it will be created using the \code{date_time} variable.
 #' @param s_start_doy \code{numerical(1)} Default `NULL` (if `NULL`, `s_start_doy = 1`. The day of year to state is the first day of year.
+#' @param drop \code{logical(1)} default `TRUE`. Whether to drop years where there are `NA` data for the rainfall.
 #' @param start_day \code{numerical(1)} The first day to calculate from in the year (1-366).
 #' @param end_day \code{numerical(1)} The last day to calculate to in the year (1-366).
 #' @param output \code{character(1)} Whether to give the start of rains by day of year (doy), date, or both. Default `"doy"`.
@@ -26,7 +27,7 @@
 #' #capacity = 60, water_balance_max = 0.5,
 #' #evaporation_value = 5)
 end_season <- function(data, date_time, station = NULL, year = NULL, rain = NULL,
-                       doy = NULL,  s_start_doy = NULL,
+                       doy = NULL,  s_start_doy = NULL, drop = TRUE,
                        start_day = 1, end_day = 366, output = c("doy", "date", "both"),
                        capacity = 60, water_balance_max = 0.5, evaporation = c("value", "variable"),
                        evaporation_value = 5, evaporation_variable = NULL){
@@ -71,7 +72,7 @@ end_season <- function(data, date_time, station = NULL, year = NULL, rain = NULL
 
   if (!is.null(station)){
     end_of_season <- data %>% 
-      dplyr::group_by(.data[[station]], .add = TRUE, .drop = FALSE) 
+      dplyr::group_by(.data[[station]], .add = TRUE, .drop = drop) 
   } else {
     end_of_season <- data
   }
@@ -84,7 +85,7 @@ end_season <- function(data, date_time, station = NULL, year = NULL, rain = NULL
     dplyr::mutate(wb_max = Reduce(f = function(x, y) pmin(pmax(x + y, 0), capacity), x = utils::tail(x=rain_max - evaporation_value, n=-1), init=0, accumulate=TRUE)) %>%
     dplyr::mutate(wb = ifelse((wb_min != wb_max) | is.na(.data[[rain]]), NA, wb_min)) %>%
     dplyr::filter(wb <= water_balance_max | is.na(.data[[rain]])) %>%
-    dplyr::group_by(.data[[year]], .add = TRUE, .drop = FALSE) %>%
+    dplyr::group_by(.data[[year]], .add = TRUE, .drop = drop) %>%
     dplyr::filter(.data[[doy]] >= start_day & .data[[doy]] <= end_day, .preserve = TRUE)
 
   if (output == "doy"){

R/get_extremes.R

@@ -40,11 +40,11 @@ get_extremes <- function(data, station = NULL, year, element, type = c("percenti
   } 
 
   if (!is.null(station)){
-    extreme_data <- extreme_data %>% dplyr::group_by(.data[[station]], .drop = FALSE)
+    extreme_data <- extreme_data %>% dplyr::group_by(.data[[station]])
   }
   extreme_data[[year]] <- factor(extreme_data[[year]])
   extreme_data <- extreme_data %>%
-    dplyr::group_by(.data[[year]], .add = TRUE, .drop = FALSE) %>%
+    dplyr::group_by(.data[[year]], .add = TRUE) %>%
     summarise(count = n())
 
   return(extreme_data)

R/probability_season_start.R

@@ -32,7 +32,7 @@ probability_season_start <- function(data, station = NULL, start_rains, doy_form
   if (!is.null(station)){
     summary_proportion <- purrr::map_df(.x = specified_day,
                                         .f =~ data %>% dplyr::mutate(LT = ifelse(.data[[start_rains]] < .x, 1, 0)) %>%
-                                          dplyr::group_by(.data[[station]], .drop = FALSE) %>%
+                                          dplyr::group_by(.data[[station]], .drop = TRUE) %>%
                                           dplyr::summarise(day = .x, proportion = sum(LT, na.rm = TRUE)/dplyr::n()))
   } else {
     summary_proportion <- purrr::map_df(.x = specified_day,

R/start_rains.R

@@ -11,6 +11,7 @@
 #' @param start_day \code{numerical(1)} The first day to calculate from in the year (1-366).
 #' @param end_day \code{numerical(1)} The last day to calculate to in the year (1-366).
 #' @param s_start_doy \code{numerical(1)} Default `NULL` (if `NULL`, `s_start_doy = 1`. The day of year to state is the first day of year.
+#' @param drop \code{logical(1)} default `TRUE`. Whether to drop years where there are `NA` data for the rainfall.
 #' @param output \code{character(1)} Whether to give the start of rains by day of year (doy), date, or both. Default `"doy"`.
 #' @param total_rainfall \code{logical(1)} default `TRUE`. Start of the rains to be defined by the total or proportion of rainfall over a period.
 #' @param over_days \code{numerical(1)} Only works if `total_rainfall = TRUE`. This is the number of days to total the rainfall over.
@@ -35,6 +36,7 @@
 #' # check against R-Instat function
 start_rains <- function(data, date_time, station = NULL, year = NULL, rain = NULL, threshold = 0.85,
                          doy = NULL, start_day = 1, end_day = 366, s_start_doy = NULL,
+                        drop = TRUE,
                          output = c("doy", "date", "both"),
                          total_rainfall = TRUE, over_days = 1, amount_rain = 20, proportion = FALSE, prob_rain_day = 0.8,
                          number_rain_days = FALSE, min_rain_days = 1, rain_day_interval = 2,
@@ -107,7 +109,7 @@ start_rains <- function(data, date_time, station = NULL, year = NULL, rain = NUL
 
   if (!is.null(station)){
     start_of_rains <- data %>% 
-      dplyr::group_by(.data[[station]], .drop = FALSE) 
+      dplyr::group_by(.data[[station]], .drop = drop) 
   } else {
     start_of_rains <- data
   }
@@ -168,7 +170,7 @@ start_rains <- function(data, date_time, station = NULL, year = NULL, rain = NUL
   }
 
   start_of_rains <- start_of_rains %>% 
-    dplyr::group_by(.data[[year]], .add = TRUE, .drop = FALSE) %>% 
+    dplyr::group_by(.data[[year]], .add = TRUE, .drop = drop) %>% 
     dplyr::filter(.data[[doy]] >= start_day & .data[[doy]] <= end_day, .preserve = TRUE)
 
   if (output == "doy"){