Skip to contents

Returns the canonical registry of the seven dimensions on which the liminal space between a business ask and a data-based solution is measured. The registry carries the code-level name used in score vectors and weights, the plain-language title and formal_name from the README, the question a business owner and a data scientist can answer for each dimension, the evidence one would collect, and the prompt_version of the template used to score it with an LLM.

Usage

liminal_dimensions()

Value

A tibble with one row per dimension and the columns name, title, formal_name, question_business, question_data_scientist, evidence and prompt_version.

Examples

liminal2::liminal_dimensions()
#> # A tibble: 7 × 7
#>   name       title formal_name question_business question_data_scient…¹ evidence
#>   <chr>      <chr> <chr>       <chr>             <chr>                  <chr>   
#> 1 shared_me… Shar… Rival oper… "When we say \"c… Do rival label defini… Label a…
#> 2 stable_gr… Stab… Outcome-ge… "Will next quart… Is the outcome-genera… Out-of-…
#> 3 convergen… Conv… Near-equal… "If two good ana… Do near-equal models … Share o…
#> 4 knowable_… Know… Target is … "Is the answer v… Is the target attaina… Skill c…
#> 5 judgeable  Judg… An externa… "Who can say, in… Is there an adjudicat… Rubric:…
#> 6 action_id… Acti… The decisi… "What will we do… Can the decision rule… Recover…
#> 7 tradeoff_… Trad… The recove… "Do the people w… Do stakeholders' loss… Consens…
#> # ℹ abbreviated name: ¹​question_data_scientist
#> # ℹ 1 more variable: prompt_version <chr>