traits.build v2.0.0 release

DESCRIPTION

@@ -1,7 +1,7 @@
 Type: Package
 Package: traits.build
 Title: A workflow for harmonising trait data from diverse sources into a documented standard structure
-Version: 1.1.0
+Version: 2.0.0
 Maintainer: Daniel Falster <[email protected]>
 Authors@R: c(
     person(given = "Daniel", family = "Falster", role = c("cre", "aut"), email = "[email protected]", comment = c(ORCID = "0000-0002-9814-092X")),
@@ -22,7 +22,8 @@ Depends:
     lubridate,
     readr,
     stringr,
-    tidyr
+    tidyr,
+    austraits
 Imports:
     crayon,
     git2r,
@@ -38,7 +39,8 @@ Imports:
     testthat,
     tibble,
     whisker,
-    yaml
+    yaml,
+    lifecycle
 Suggests:
     furrr,
     remake,
@@ -57,8 +59,10 @@ Suggests:
     zip,
     covr
 Remotes:
-    richfitz/remake
+    richfitz/remake,
+    traitecoevo/austraits
 Encoding: UTF-8
 VignetteBuilder: knitr
-RoxygenNote: 7.2.3
+RoxygenNote: 7.3.2
 Config/testthat/edition: 3
+Roxygen: list(markdown = TRUE)

NAMESPACE

@@ -1,19 +1,22 @@
 # Generated by roxygen2: do not edit by hand
 
 export("%>%")
+export(bind_databases)
 export(build_add_version)
 export(build_combine)
 export(build_setup_pipeline)
 export(check_pivot_duplicates)
+export(convert_df_to_list)
+export(convert_list_to_df1)
+export(convert_list_to_df2)
 export(dataset_build)
 export(dataset_configure)
 export(dataset_find_taxon)
 export(dataset_process)
 export(dataset_report)
 export(dataset_test)
 export(dataset_update_taxonomy)
-export(db_traits_pivot_longer)
-export(db_traits_pivot_wider)
+export(flatten_database)
 export(get_schema)
 export(get_unit_conversions)
 export(metadata_add_contexts)
@@ -32,7 +35,6 @@ export(metadata_exclude_observations)
 export(metadata_find_taxonomic_change)
 export(metadata_remove_taxonomic_change)
 export(metadata_update_taxonomic_change)
-export(plot_trait_distribution_beeswarm)
 export(read_csv_char)
 export(read_metadata)
 export(util_df_to_list)
@@ -44,6 +46,11 @@ export(util_list_to_df1)
 export(util_list_to_df2)
 export(write_metadata)
 export(write_plaintext)
+importFrom(austraits,bind_databases)
+importFrom(austraits,convert_df_to_list)
+importFrom(austraits,convert_list_to_df1)
+importFrom(austraits,convert_list_to_df2)
+importFrom(austraits,flatten_database)
 importFrom(crayon,"%+%")
 importFrom(crayon,blue)
 importFrom(crayon,green)
@@ -59,6 +66,7 @@ importFrom(dplyr,filter)
 importFrom(dplyr,full_join)
 importFrom(dplyr,mutate)
 importFrom(dplyr,select)
+importFrom(lifecycle,deprecated)
 importFrom(lubridate,dmy)
 importFrom(magrittr,"%>%")
 importFrom(purrr,map_chr)
@@ -73,9 +81,7 @@ importFrom(stringr,str_remove_all)
 importFrom(stringr,str_replace_all)
 importFrom(stringr,str_squish)
 importFrom(styler,style_text)
-importFrom(testthat,compare)
 importFrom(testthat,context)
-importFrom(testthat,expect)
 importFrom(testthat,expect_silent)
 importFrom(testthat,local_edition)
 importFrom(testthat,test_that)

NEWS.md

@@ -1,3 +1,21 @@
+# traits.build 2.0.0
+
+- traits.build paper published in Sep 2024 in Ecological Informatics (DOI: [10.1016/j.ecoinf.2024.102773](https://doi.org/10.1016/j.ecoinf.2024.102773))
+- Added standard error and standard deviation as value types
+- Moved functions to austraits package and made austraits package a dependancy
+- Renamed some of the functions that are now moved to austraits package 
+    * `bind_databases` <-- `build_combine`
+    * `convert_df_to_list` <-- `util_df_to_list`
+    * `convert_list_to_df1` <-- `util_list_to_df1`
+    * `convert_list_to_df2` <-- `util_list_to_df2`
+- Renamed functions still *also* assigned their old name, with a deprecation warning indicating the new name  
+- `plot_trait_distribution_beeswarm`, `trait_pivot_longer` and `trait_pivot_wider` had been in both austraits and traits.build packages and have now been removed from traits.build
+- Import new austraits function `flatten_database` (had been suggested to be `database_create_combined_table`)
+- Refactoring of test functions used by `dataset_test`
+- Added tests using the `dataset_test` function, so it is checked explicitly by traits.build (run on Example datasets)
+- Minor bug fixes
+- Minor updates to ontology (now version 1.0)
+
 # traits.build 1.1.0
 
 - Small bugfixes in dataset_test

R/pivot.R

@@ -30,108 +30,3 @@ check_pivot_wider <- function(dataset) {
   }
 
 }
-
-
-#' @title Pivot long format data into a wide format
-#'
-#' @description `trait_pivot_wider` "widens" long format data ("tidy data").
-#'
-#' Databases built with `traits.build` are organised in a long format where observations are on different rows and the
-#' type of observation is denoted by various identifying columns (e.g `trait_name`, `dataset_id`,
-#' `observation_id`, etc.).
-#' This function converts the data into wide format so that each trait in its own column.
-#'
-#' @param traits The traits table from database (list object)
-#' @return A tibble in wide format
-#' @details `trait_pivot_wider` will return a single wide tibble; note that some meta-data columns
-#' (unit, replicates, measurement_remarks, basis_of_record, basis_of_value) will be excluded to
-#' produce a useful wide tibble.
-#' @examples
-#' \dontrun{
-#' data <- austraits$traits %>% filter(dataset_id == "Falster_2003")
-#' data # Long format
-#' traits_wide <- trait_pivot_wider(data)
-#' traits_wide # Wide format
-#' }
-#' @author Daniel Falster - [email protected]
-#' @export
-db_traits_pivot_wider <- function(traits) {
-
-  metadata_cols <- c("unit", "replicates", "measurement_remarks", "basis_of_value")
-
-  # A check for if there are more than 1 value_type for a given taxon_name, observation_id and method
-  check_value_type <- traits %>%
-    dplyr::select(dplyr::all_of(c(
-      "trait_name", "value", "dataset_id", "observation_id", "method_id", "method_context_id",
-      "repeat_measurements_id", "value_type"))) %>%
-    dplyr::group_by(
-      .data$dataset_id, .data$observation_id, .data$method_id,
-      .data$method_context_id, .data$repeat_measurements_id) %>%
-    dplyr::summarise(n_value_type = length(unique(.data$value_type))) %>%
-    dplyr::arrange(.data$observation_id) %>%
-    dplyr::filter(.data$n_value_type > 1)
-
-  if (nrow(check_value_type) > 1) {
-
-    traits %>%
-      tidyr::pivot_wider(
-        names_from = "trait_name",
-        values_from = "value",
-        id_cols = -dplyr::all_of(metadata_cols)
-      )
-
-  } else {
-
-    metadata_cols <- c(metadata_cols, "value_type")
-
-    traits %>%
-      tidyr::pivot_wider(
-        names_from = "trait_name",
-        values_from = "value",
-        id_cols = -dplyr::all_of(metadata_cols)
-      )
-  }
-
-}
-
-
-#' @title Pivot wide format data into a long format
-#'
-#' @description `trait_pivot_longer` "gathers" wide format data into a "tidy" format.
-#'
-#' This function converts the data into long format where observations are on different rows and the type of
-#' observation is denoted by the `trait_name` column.
-#' In other words, `trait_pivot_longer` reverts the actions of `trait_pivot_wider`.
-#' @param wide_data Output from `trait_pivot_wider` (a tibble of wide data)
-#' @return A tibble in long format
-#' @details
-#' `trait_pivot_longer` will return a tibble with fewer columns than the original traits table
-#' The excluded columns include: "unit", "replicates", "measurement_remarks", "basis_of_record",
-#' "basis_of_value" # Double check #TODO
-#'
-#' @examples
-#' \dontrun{
-#' data <- austraits$traits %>%
-#' filter(dataset_id == "Falster_2003")
-#' data # Long format
-#' traits_wide <- trait_pivot_wider(data)
-#' traits_wide # Wide format
-#'
-#' values_long <- trait_pivot_longer(traits_wide)
-#' }
-#' @author Daniel Falster - [email protected]
-#' @author Fonti Kar - [email protected]
-#' @export
-db_traits_pivot_longer <- function(wide_data) {
-
-  # The start of the trait columns is after `original_name`
-  start_of_trait_cols <- which(names(wide_data) == "original_name") + 1
-
-  wide_data %>%
-    tidyr::pivot_longer(
-      cols = start_of_trait_cols:ncol(.),
-      names_to = "trait_name",
-      values_drop_na = TRUE
-    )
-
-}