Match model output data to their model tasks in config_tasks. — match_tbl_to_model

Useful for performing model task specific checks on model output.

Usage

match_tbl_to_model_task(
  tbl,
  config_tasks,
  round_id,
  output_types = NULL,
  derived_task_ids = get_config_derived_task_ids(config_tasks, round_id),
  order_by_config = FALSE
)

Arguments

tbl

a tibble/data.frame of the contents of the file being validated. Column types must all be character: the config's values are converted to character when they are extracted, and are compared against this table without further conversion. Every task ID column the round defines must be present.

config_tasks

a list version of the content's of a hub's tasks.json config file, accessed through the "config_tasks" attribute of a <hub_connection> object or function hubUtils::read_config().

round_id

Character string. Round identifier. If the round is set to round_id_from_variable: true, IDs are values of the task ID defined in the round's round_id property of config_tasks. Otherwise should match round's round_id value in config. Ignored if hub contains only a single round.

output_types

Character vector of output type names to include. Use to subset for grids for specific output types.

derived_task_ids

Character vector of derived task ID names, or NULL for none. A derived task ID's value follows from the values of other task IDs. A derived task ID cannot therefore further distinguish a row beyond the values of the task IDs it is derived from. Derived task ID columns are skipped, and returned unchanged.

order_by_config

Logical. How to order each modeling task's rows. FALSE, the default, leaves them in the order they were submitted in. TRUE sorts them into the order the config lists their values in: on output_type first, so rows of one output type sit together, then on output_type_id, so they ascend within each, then on the task IDs to break ties.

What is sorted on is each value's position in the config, not the value itself. pmf categories show why that matters: "low", "moderate" and "high" have no useful alphabetical order, but the config lists them in the order they belong in. check_tbl_value_col_ascending() asks for this order, because it reads values in the order the rows arrive in.

In the config a task ID's values are split into required and optional. Extracting them collapses the two into a single order, required values first, then optional ones, and that is the order sorted on.

Sample rows are ordered by their task ID values only.

Value

A list with one element per model task in the round, each a tbl_df of the model output rows matched to that model task. A model task that offers none of the requested output_types gets NULL. Rows that match no model task are not returned.

Details

Sample output_type_id values are returned as submitted. The submitter chooses them, so this function does not check them against the config.

Examples

hub_path <- system.file("testhubs/samples", package = "hubValidations")
tbl <- read_model_out_file(
  file_path = "flu-base/2022-10-22-flu-base.csv",
  hub_path, coerce_types = "chr"
)
config_tasks <- read_config(hub_path, "tasks")
match_tbl_to_model_task(tbl, config_tasks, round_id = "2022-10-22")
#> [[1]]
#> # A tibble: 60 × 8
#>    reference_date target            horizon location target_end_date output_type
#>    <chr>          <chr>             <chr>   <chr>    <chr>           <chr>      
#>  1 2022-10-22     wk flu hosp rate… 0       01       2022-10-22      pmf        
#>  2 2022-10-22     wk flu hosp rate… 0       01       2022-10-22      pmf        
#>  3 2022-10-22     wk flu hosp rate… 0       01       2022-10-22      pmf        
#>  4 2022-10-22     wk flu hosp rate… 0       01       2022-10-22      pmf        
#>  5 2022-10-22     wk flu hosp rate… 1       01       2022-10-29      pmf        
#>  6 2022-10-22     wk flu hosp rate… 1       01       2022-10-29      pmf        
#>  7 2022-10-22     wk flu hosp rate… 1       01       2022-10-29      pmf        
#>  8 2022-10-22     wk flu hosp rate… 1       01       2022-10-29      pmf        
#>  9 2022-10-22     wk flu hosp rate… 2       01       2022-11-05      pmf        
#> 10 2022-10-22     wk flu hosp rate… 2       01       2022-11-05      pmf        
#> # ℹ 50 more rows
#> # ℹ 2 more variables: output_type_id <chr>, value <chr>
#> 
#> [[2]]
#> # A tibble: 1,530 × 8
#>    reference_date target          horizon location target_end_date output_type
#>    <chr>          <chr>           <chr>   <chr>    <chr>           <chr>      
#>  1 2022-10-22     wk inc flu hosp 0       US       2022-10-22      median     
#>  2 2022-10-22     wk inc flu hosp 1       US       2022-10-29      median     
#>  3 2022-10-22     wk inc flu hosp 2       US       2022-11-05      median     
#>  4 2022-10-22     wk inc flu hosp 0       01       2022-10-22      median     
#>  5 2022-10-22     wk inc flu hosp 1       01       2022-10-29      median     
#>  6 2022-10-22     wk inc flu hosp 2       01       2022-11-05      median     
#>  7 2022-10-22     wk inc flu hosp 0       02       2022-10-22      median     
#>  8 2022-10-22     wk inc flu hosp 1       02       2022-10-29      median     
#>  9 2022-10-22     wk inc flu hosp 2       02       2022-11-05      median     
#> 10 2022-10-22     wk inc flu hosp 0       04       2022-10-22      median     
#> # ℹ 1,520 more rows
#> # ℹ 2 more variables: output_type_id <chr>, value <chr>
#> 
match_tbl_to_model_task(tbl, config_tasks,
  round_id = "2022-10-22",
  output_types = "sample"
)
#> [[1]]
#> NULL
#> 
#> [[2]]
#> # A tibble: 1,500 × 8
#>    reference_date target          horizon location target_end_date output_type
#>    <chr>          <chr>           <chr>   <chr>    <chr>           <chr>      
#>  1 2022-10-22     wk inc flu hosp 0       01       2022-10-22      sample     
#>  2 2022-10-22     wk inc flu hosp 0       01       2022-10-22      sample     
#>  3 2022-10-22     wk inc flu hosp 0       01       2022-10-22      sample     
#>  4 2022-10-22     wk inc flu hosp 0       01       2022-10-22      sample     
#>  5 2022-10-22     wk inc flu hosp 0       01       2022-10-22      sample     
#>  6 2022-10-22     wk inc flu hosp 0       01       2022-10-22      sample     
#>  7 2022-10-22     wk inc flu hosp 0       01       2022-10-22      sample     
#>  8 2022-10-22     wk inc flu hosp 0       01       2022-10-22      sample     
#>  9 2022-10-22     wk inc flu hosp 0       01       2022-10-22      sample     
#> 10 2022-10-22     wk inc flu hosp 0       01       2022-10-22      sample     
#> # ℹ 1,490 more rows
#> # ℹ 2 more variables: output_type_id <chr>, value <chr>
#>