Merge pull request nationalparkservice#110 from RobLBaker/master

minor updates and fixes

NAMESPACE

@@ -20,7 +20,7 @@ export(get_taxon_rank)
 export(get_utm_zone)
 export(long2UTM)
 export(order_cols)
-export(removeEmptyTables)
+export(remove_empty_tables)
 export(replace_blanks)
 export(te_check)
 export(utm_to_ll)

NEWS.md

@@ -1,3 +1,10 @@
+# QCkit v0.1.8 (not yet released)
+
+2024-06-27
+* bug fixes for `generate_ll_from_utm()`
+* add function `remove_empty_tables()` (and associated unit tests)
+* update documentation for `replace blanks()` to indicate it can replace blanks with more than just NA
+
 # QCkit v0.1.7
 2024-05-08
 * Updated the `replace_blanks()` function to accept any missing value code a user inputs (but it still defaults to NA).

R/geography.R

@@ -425,7 +425,11 @@ generate_ll_from_utm <- function(df,
 
   # Separate df with just coordinates. We'll filter out any NA rows.
   coord_df <- df %>%
-    dplyr::select(`_UTMJOINCOL`, {{EastingCol}}, {{NorthingCol}}, {{ZoneCol}}, {{DatumCol}})
+    dplyr::select(`_UTMJOINCOL`,
+                  {{EastingCol}},
+                  {{NorthingCol}},
+                  {{ZoneCol}},
+                  {{DatumCol}})
 
   withr::with_envvar(c("PROJ_LIB" = ""), {  # This is a fix for the proj library bug in R (see pinned post "sf::st_read() of geojson not getting CRS" in IMData General Discussion).
       coord_df <- coord_df %>%
@@ -437,14 +441,19 @@ generate_ll_from_utm <- function(df,
 
     na_row_count <- nrow(df) - nrow(coord_df)
     if (na_row_count > 0) {
-      warning(paste(na_row_count, "rows are missing UTM coordinates, zone, and/or datum information."), call. = FALSE)
+      warning(paste(
+        na_row_count,
+        "rows are missing UTM coordinates, zone, and/or datum information."),
+        call. = FALSE)
     }
 
     ## Set up CRS for lat/long data
-    latlong_CRS <- sp::CRS(glue::glue("+proj=longlat +datum={latlong_datum}"))  # CRS for our new lat/long values
+    latlong_CRS <- sp::CRS(glue::glue("+proj=longlat +datum={latlong_datum}"))
+    # CRS for our new lat/long values
 
     # Loop through each datum and zone in the data
-    zones_datums <- dplyr::select(coord_df, {{ZoneCol}}, {{DatumCol}}) %>%  # Get vector of zones present in data
+    zones_datums <- dplyr::select(coord_df, {{ZoneCol}}, {{DatumCol}}) %>%
+      # Get vector of zones present in data
       unique()
 
     new_coords <- sapply(1:nrow(zones_datums), function(zone_datum_index) {
@@ -454,7 +463,8 @@ generate_ll_from_utm <- function(df,
         zone_num <- current_zone
         north_south <- ""
       } else {
-        zone_num <- stringr::str_extract(current_zone, "\\d+")  # sp::CRS wants zone number only, e.g. 11, not 11N
+        zone_num <- stringr::str_extract(current_zone, "\\d+")
+        # sp::CRS wants zone number only, e.g. 11, not 11N
         zone_letter <- tolower(stringr::str_extract(current_zone, "[A-Za-z]"))
         if (!is.na(zone_letter) && zone_letter == "s") {
           north_south <- " +south"
@@ -469,17 +479,21 @@ generate_ll_from_utm <- function(df,
       filtered_df <- coord_df %>%
         dplyr::filter((!!rlang::ensym(ZoneCol) == current_zone & !!rlang::ensym(DatumCol) == current_datum))
       sp_utm <- sp::SpatialPoints(filtered_df %>%
-                                    dplyr::select({{EastingCol}}, {{NorthingCol}}) %>%
+                                    dplyr::select({{EastingCol}},
+                                                  {{NorthingCol}}) %>%
                                     as.matrix(),
-                                  proj4string = utm_CRS)  # Convert UTM columns into a SpatialPoints object
-      sp_geo <- sp::spTransform(sp_utm, latlong_CRS) %>%  # Transform UTM to Lat/Long
+                                  proj4string = utm_CRS)
+      # Convert UTM columns into a SpatialPoints object
+      sp_geo <- sp::spTransform(sp_utm, latlong_CRS) %>%
+        # Transform UTM to Lat/Long
         tibble::as_tibble()
 
       # Add lat/long columns back into the original dataframe
       latlong <- tibble::tibble(`_UTMJOINCOL` = filtered_df$`_UTMJOINCOL`,
                                 decimalLatitude = sp_geo[[2]],
                                 decimalLongitude = sp_geo[[1]],
-                                LatLong_CRS = latlong_CRS@projargs)  # Store the coordinate reference system PROJ string in the dataframe
+                                LatLong_CRS = latlong_CRS@projargs)
+      # Store the coordinate reference system PROJ string in the dataframe
 
 
       return(latlong)

R/remove_empty_tables.R

@@ -11,9 +11,9 @@
 #'                   item_b = mtcars,
 #'                   item_c = iris)
 #'
-#' tidy_list <- removeEmptyTables(test_list)
+#' tidy_list <- remove_empty_tables(test_list)
 #'
-removeEmptyTables <- function(df_list) {
+remove_empty_tables <- function(df_list) {
   non_empty_list <- purrr::compact(df_list)  # Remove empty dataframes
   tables_removed <- setdiff(names(df_list), names(non_empty_list))  # Get names of removed dataframes
 

R/replace_blanks.R

@@ -1,7 +1,8 @@
-#' Replaces all blank cells with NA
+#' Replaces all blank cells a missing value code of your choice
 #'
-#' @details `replace_blanks()` is particularly useful for exporting data from a
-#' database (such as access) and converting it to a data package with metadata.
+#' @description `replace_blanks()` is particularly useful for exporting data
+#' from a database (such as access) and converting it to a data package with
+#' metadata.
 #'
 #' `replace_blanks()` will import all .csv files in the specified working
 #' directory. The files are then written back out to the same directory,
@@ -10,9 +11,23 @@
 #' If no missing value is specified, the function defaults to replacing all
 #' blanks with "NA".
 #'
-#' One exception is if a .csv contains NO data (i.e. just column names and no
-#' data in any of the cells). In this case, the blanks will not be replaced with
-#' NA (as the function cannot determine how many NAs to include).
+#' Please keep in mind the "missing" is a general term for all data
+#' not present in the data file or data package. Although you may have a very
+#' good reason for not providing data and that data may not, from the data
+#' package creator's perspective, be "missing" (maybe you never intended to
+#' collect it) from a data package user's perspective any data that is not in
+#' the data package is effectively "missing" from the data package. Therefore,
+#' it is critical to document in metadata any data that are absent with an
+#' appropriate "missingValueCode" and "missingValueDefinition". These terms are
+#' defined by the metadata schema and are broadly used to apply to any data not
+#' present.
+#'
+#' This function will replace all empty cells and all cells with NA with a
+#' "missingValueCode" of your choice (although it defaults to NA).
+#'
+#' @details One exception is if a .csv contains NO data (i.e. just column names
+#' and no data in any of the cells). In this case, the blanks will not be
+#' replaced with NA (as the function cannot determine how many NAs to include).
 #'
 #' @param directory String. Path to the file(s) to have blanks replaced with
 #' NAs. Defaults to the working directory of the project (here::here())