Merge pull request #157 from hubverse-org/ak/import-get-hub-fns

Import/re-export functions
hubverse-org · Nov 13, 2024 · 9a3101e · 9a3101e
2 parents d978fdd + fa2bad2
commit 9a3101e
Show file tree

Hide file tree

Showing 25 changed files with 108 additions and 131 deletions.
diff --git a/DESCRIPTION b/DESCRIPTION
@@ -40,13 +40,12 @@ Imports:
     fs,
     gh,
     hubAdmin (>= 1.0.0),
-    hubData (>= 1.1.0),
-    hubUtils (>= 0.1.2),
+    hubData (>= 1.2.3.9000),
+    hubUtils (>= 0.2.0.9000),
     jsonlite,
     jsonvalidate,
     lifecycle,
     lubridate,
-    magrittr,
     purrr,
     rlang,
     stringr,

diff --git a/NAMESPACE b/NAMESPACE
@@ -41,8 +41,10 @@ export(check_tbl_values)
 export(check_tbl_values_required)
 export(check_valid_round_id)
 export(check_valid_round_id_col)
+export(coerce_to_hub_schema)
 export(combine)
 export(create_custom_check)
+export(create_hub_schema)
 export(expand_model_out_grid)
 export(get_tbl_compound_taskid_set)
 export(is_any_error)
@@ -60,6 +62,8 @@ export(opt_check_tbl_col_timediff)
 export(opt_check_tbl_counts_lt_popn)
 export(opt_check_tbl_horizon_timediff)
 export(parse_file_name)
+export(read_config)
+export(read_config_file)
 export(read_model_out_file)
 export(submission_tmpl)
 export(try_check)
@@ -69,6 +73,14 @@ export(validate_model_metadata)
 export(validate_pr)
 export(validate_submission)
 export(validate_submission_time)
+importFrom(dplyr,"%>%")
+importFrom(hubData,coerce_to_hub_schema)
+importFrom(hubData,create_hub_schema)
+importFrom(hubUtils,get_derived_task_ids)
+importFrom(hubUtils,get_hub_file_formats)
+importFrom(hubUtils,get_hub_model_output_dir)
+importFrom(hubUtils,get_hub_timezone)
+importFrom(hubUtils,read_config)
+importFrom(hubUtils,read_config_file)
 importFrom(lubridate,"%within%")
-importFrom(magrittr,"%>%")
 importFrom(rlang,"!!!")
diff --git a/NEWS.md b/NEWS.md
@@ -1,5 +1,9 @@
 # hubValidations (development version)
 
+* Re-exported functions useful for modelers (#149):
+  - `hubUtils::read_config()` and `hubUtils::read_config_file()` for reading in hub configuration files.
+  - `hubData::create_hub_schema()` and `hubData::coerce_to_hub_schema()` for creating and coercing data to the hub schema.
+
 # hubValidations 0.8.0
 
 * Custom checks no longer fail if validation is run outside of the root of the hub (#141)

diff --git a/R/imports-hubUtils.R b/R/imports-hubUtils.R
@@ -0,0 +1,11 @@
+#' @importFrom hubUtils get_hub_file_formats
+NULL
+
+#' @importFrom hubUtils get_hub_timezone
+NULL
+
+#' @importFrom hubUtils get_hub_model_output_dir
+NULL
+
+#' @importFrom hubUtils get_derived_task_ids
+NULL
diff --git a/R/utils-pipe.R → R/reexport-dplyr.R b/R/utils-pipe.R → R/reexport-dplyr.R
@@ -1,13 +1,10 @@
 #' Pipe operator
 #'
-#' See \code{magrittr::\link[magrittr:pipe]{\%>\%}} for details.
-#'
-#' @name %>%
 #' @rdname pipe
 #' @keywords internal
+#' @name  %>%
+#' @importFrom dplyr %>%
 #' @export
-#' @importFrom magrittr %>%
-#' @usage lhs \%>\% rhs
 #' @param lhs A value or the magrittr placeholder.
 #' @param rhs A function call using the magrittr semantics.
 #' @return The result of calling `rhs(lhs)`.

diff --git a/R/reexport-hubData.R b/R/reexport-hubData.R
@@ -0,0 +1,7 @@
+#' @importFrom hubData create_hub_schema
+#' @export
+hubData::create_hub_schema
+
+#' @importFrom hubData coerce_to_hub_schema
+#' @export
+hubData::coerce_to_hub_schema
diff --git a/R/reexport-hubUtils.R b/R/reexport-hubUtils.R
@@ -0,0 +1,7 @@
+#' @importFrom hubUtils read_config
+#' @export
+hubUtils::read_config
+
+#' @importFrom hubUtils read_config_file
+#' @export
+hubUtils::read_config_file
diff --git a/R/utils.R b/R/utils.R
@@ -1,26 +1,3 @@
-get_hub_file_formats <- function(hub_path, round_id) {
-  config_tasks <- hubUtils::read_config(hub_path, "tasks")
-  round_idx <- hubUtils::get_round_idx(config_tasks, round_id)
-  file_formats <- config_tasks[["rounds"]][[round_idx]]$file_format
-  if (!is.null(file_formats)) {
-    return(file_formats)
-  }
-
-  config_admin <- hubUtils::read_config(hub_path, "admin")
-  config_admin$file_format
-}
-
-get_hub_timezone <- function(hub_path) {
-  config_admin <- hubUtils::read_config(hub_path, "admin")
-  config_admin$timezone
-}
-
-get_hub_model_output_dir <- function(hub_path) {
-  config_admin <- hubUtils::read_config(hub_path, "admin")
-  model_output_dir <- config_admin$model_output_dir
-  if (is.null(model_output_dir)) "model-output" else model_output_dir
-}
-
 get_file_target_metadata <- function(hub_path, file_path) {
   round_config <- get_file_round_config(file_path, hub_path)
 

diff --git a/man/check_tbl_col_types.Rd b/man/check_tbl_col_types.Rd
diff --git a/man/expand_model_out_grid.Rd b/man/expand_model_out_grid.Rd
diff --git a/man/opt_check_tbl_col_timediff.Rd b/man/opt_check_tbl_col_timediff.Rd
diff --git a/man/opt_check_tbl_horizon_timediff.Rd b/man/opt_check_tbl_horizon_timediff.Rd
diff --git a/man/pipe.Rd b/man/pipe.Rd
diff --git a/man/read_model_out_file.Rd b/man/read_model_out_file.Rd
diff --git a/man/reexports.Rd b/man/reexports.Rd
diff --git a/man/validate_model_data.Rd b/man/validate_model_data.Rd
diff --git a/man/validate_pr.Rd b/man/validate_pr.Rd
diff --git a/man/validate_submission.Rd b/man/validate_submission.Rd
diff --git a/tests/testthat/_snaps/check_tbl_value_col.md b/tests/testthat/_snaps/check_tbl_value_col.md
@@ -12,9 +12,9 @@
     Code
       check_tbl_value_col(tbl, round_id, file_path, hub_path)
     Output
-      <error/check_failure>
-      Error:
-      ! Values in column `value` are not all valid with respect to modeling task config.  Values 196.83, 367.89, 244.85, 427.975, 310.9, 499.9, 394.8, 461.85, 629.7, 534.6, 599.75, 727.5, 669.7, 725.562456357284, 801.689162048122, 800.239938093011, 915.5, 949.59, ..., 2775.56600111008, and 2139.07 cannot be coerced to expected data type "integer" for output type "quantile".
+      <message/check_success>
+      Message:
+      Values in column `value` all valid with respect to modeling task config.
 
 # check_tbl_value_col errors correctly
 
@@ -32,7 +32,7 @@
     Output
       <error/check_failure>
       Error:
-      ! Values in column `value` are not all valid with respect to modeling task config.  Value "fail" cannot be coerced to expected data type "integer" for output type "quantile".Values in column `value` are not all valid with respect to modeling task config.  Values "196.83", "244.85", "310.9", "499.9", "394.8", "461.85", "629.7", "534.6", "599.75", "727.5", "669.7", "725.562456357284", "801.689162048122", "800.239938093011", "915.5", "949.59", "938.67", "962.03", ..., "2775.56600111008", and "2139.07" cannot be coerced to expected data type "integer" for output type "quantile".Values in column `value` are not all valid with respect to modeling task config.  Value -6 is smaller than allowed minimum value 0 for output type "quantile".
+      ! Values in column `value` are not all valid with respect to modeling task config.  Value "fail" cannot be coerced to expected data type "integer" for output type "quantile".Values in column `value` are not all valid with respect to modeling task config.  Value -6 is smaller than allowed minimum value 0 for output type "quantile".
 
 # Ignoring derived_task_ids in check_tbl_value_col works
 

diff --git a/tests/testthat/_snaps/check_tbl_value_col_sum1.md b/tests/testthat/_snaps/check_tbl_value_col_sum1.md
@@ -16,7 +16,7 @@
        $ message       : chr "Values in `value` column do not sum to 1 for all unique task ID value combination of pmf\n    output types. \n "| __truncated__
        $ trace         : NULL
        $ parent        : NULL
-       $ where         : chr "umass_ens/2023-05-08-umass_ens.csv"
+       $ where         : chr "umass-ens/2023-05-08-umass-ens.csv"
        $ error_tbl     : tibble [1 x 5] (S3: tbl_df/tbl/data.frame)
         ..$ forecast_date: Date[1:1], format: "2023-05-08"
         ..$ horizon      : int 2
@@ -36,7 +36,7 @@
        $ message       : chr "Values in `value` column do not sum to 1 for all unique task ID value combination of pmf\n    output types. \n "| __truncated__
        $ trace         : NULL
        $ parent        : NULL
-       $ where         : chr "umass_ens/2023-05-08-umass_ens.csv"
+       $ where         : chr "umass-ens/2023-05-08-umass-ens.csv"
        $ error_tbl     : tibble [1 x 5] (S3: tbl_df/tbl/data.frame)
         ..$ forecast_date: Date[1:1], format: "2023-05-08"
         ..$ horizon      : int 2
@@ -54,7 +54,7 @@
     Output
       List of 5
        $ message       : chr "Values in `value` column do sum to 1 for all unique task ID value combination of pmf\n    output types. \n "
-       $ where         : chr "umass_ens/2023-05-08-umass_ens.csv"
+       $ where         : chr "umass-ens/2023-05-08-umass-ens.csv"
        $ error_tbl     : NULL
        $ call          : chr "check_tbl_value_col_sum1"
        $ use_cli_format: logi TRUE

diff --git a/tests/testthat/_snaps/utils.md b/tests/testthat/_snaps/utils.md
@@ -1,40 +1,3 @@
-# get_hub_file_formats works
-
-    Code
-      get_hub_file_formats(hub_path = system.file("testhubs/simple", package = "hubValidations"),
-      round_id = "2022-10-08")
-    Output
-      [1] "csv"     "parquet" "arrow"  
-
----
-
-    Code
-      get_hub_file_formats(hub_path = system.file("testhubs/flusight", package = "hubUtils"),
-      round_id = "2023-01-30")
-    Output
-      [1] "csv"     "parquet" "arrow"  
-
-# get_hub_timezone works
-
-    Code
-      get_hub_timezone(hub_path = system.file("testhubs/simple", package = "hubValidations"))
-    Output
-      [1] "US/Eastern"
-
-# get_hub_model_output_dir works
-
-    Code
-      get_hub_model_output_dir(hub_path = system.file("testhubs/simple", package = "hubValidations"))
-    Output
-      [1] "model-output"
-
----
-
-    Code
-      get_hub_model_output_dir(hub_path = system.file("testhubs/flusight", package = "hubUtils"))
-    Output
-      [1] "forecasts"
-
 # get_file_round_id works
 
     Code

diff --git a/tests/testthat/_snaps/validate_model_data.md b/tests/testthat/_snaps/validate_model_data.md
@@ -167,14 +167,12 @@
         ..$ call          : chr "check_tbl_colnames"
         ..$ use_cli_format: logi TRUE
         ..- attr(*, "class")= chr [1:5] "check_success" "hub_check" "rlang_message" "message" ...
-       $ col_types         :List of 6
-        ..$ message       : chr "Column data types do not match hub schema. \n `output_type_id` should be \"character\" not \"double\"."
-        ..$ trace         : NULL
-        ..$ parent        : NULL
+       $ col_types         :List of 4
+        ..$ message       : chr "Column data types match hub schema. \n "
         ..$ where         : chr "hub-ensemble/2023-05-08-hub-ensemble.parquet"
         ..$ call          : chr "check_tbl_col_types"
         ..$ use_cli_format: logi TRUE
-        ..- attr(*, "class")= chr [1:5] "check_failure" "hub_check" "rlang_error" "error" ...
+        ..- attr(*, "class")= chr [1:5] "check_success" "hub_check" "rlang_message" "message" ...
        $ valid_vals        :List of 5
         ..$ message       : chr "`tbl` contains valid values/value combinations.  \n "
         ..$ where         : chr "hub-ensemble/2023-05-08-hub-ensemble.parquet"
@@ -201,14 +199,12 @@
         ..$ call          : chr "check_tbl_values_required"
         ..$ use_cli_format: logi TRUE
         ..- attr(*, "class")= chr [1:5] "check_success" "hub_check" "rlang_message" "message" ...
-       $ value_col_valid   :List of 6
-        ..$ message       : chr "Values in column `value` are not all valid with respect to modeling task config. \n  Values 196.83, 367.89, 244"| __truncated__
-        ..$ trace         : NULL
-        ..$ parent        : NULL
+       $ value_col_valid   :List of 4
+        ..$ message       : chr "Values in column `value` all valid with respect to modeling task config. \n "
         ..$ where         : chr "hub-ensemble/2023-05-08-hub-ensemble.parquet"
         ..$ call          : chr "check_tbl_value_col"
         ..$ use_cli_format: logi TRUE
-        ..- attr(*, "class")= chr [1:5] "check_failure" "hub_check" "rlang_error" "error" ...
+        ..- attr(*, "class")= chr [1:5] "check_success" "hub_check" "rlang_message" "message" ...
        $ value_col_non_desc:List of 5
         ..$ message       : chr "Values in `value` column are non-decreasing as output_type_ids increase for all unique task ID\n    value/outpu"| __truncated__
         ..$ where         : chr "hub-ensemble/2023-05-08-hub-ensemble.parquet"

diff --git a/tests/testthat/test-check_tbl_value_col_sum1.R b/tests/testthat/test-check_tbl_value_col_sum1.R
@@ -1,6 +1,6 @@
 test_that("check_tbl_value_col_sum1 works", {
   hub_path <- system.file("testhubs/flusight", package = "hubUtils")
-  file_path <- "umass_ens/2023-05-08-umass_ens.csv"
+  file_path <- "umass-ens/2023-05-08-umass-ens.csv"
   round_id <- "2023-05-08"
   tbl <- hubValidations::read_model_out_file(file_path, hub_path)
   expect_snapshot(
@@ -10,7 +10,7 @@ test_that("check_tbl_value_col_sum1 works", {
 
 test_that("check_tbl_value_col_sum1 errors correctly", {
   hub_path <- system.file("testhubs/flusight", package = "hubUtils")
-  file_path <- "umass_ens/2023-05-08-umass_ens.csv"
+  file_path <- "umass-ens/2023-05-08-umass-ens.csv"
   round_id <- "2023-05-08"
   tbl <- hubValidations::read_model_out_file(file_path, hub_path)
   tbl$value[1] <- 0.1