The Observation Theory EncyclopediaFrom TSKAboutBy kindBy chapterBy Lean fileLedgerProvenance

Kendall correlation concept

DefinitionThe fraction of item pairs two rankings order the same way, minus the fraction they order differently. Equation 0.15. Also Kendall, rank correlation, rank agreement.
ExampleRankings (1, 2, 3) and (1, 3, 2) agree on two pairs and disagree on one, Kendall tau one third.
BookData Mining as Observation, draft 0.2, commit f3914f0; entry id kendall-correlation, kind concept.
Statusno ledger row names this entry. Corrections: none recorded.
Defining equation

Book equation 0.15.

Assumptions and scope
  • The fraction of item pairs two rankings order the same way minus the fraction they order differently. It lies between minus one and one, it is one when no pair disagrees, and since it depends only on the orderings a strictly increasing transform of either score leaves it unchanged.
  • It is the rank-agreement statistic the rank certificate bounds from below, and a floor on it is a floor on the consumer’s rankings, not on any distance.
Prior artnone recorded
Evidencelean/DataMiningAsObservation/Kendall.lean
Reviewednot yet reviewed; generated 2026-09-10 from records at the commits on the provenance page.
rank of one score against rank of another
Concordant pairs less discordant pairs over all pairs.

Equation

Book equation 0.15.

\[\tau=\frac{\#\{\text{concordant pairs}\}-\#\{\text{discordant pairs}\}}{n(n-1)/2}.\]

Conditions

Conditions are curated in entries.toml rather than read from a record.

Ledger

none

First stated

Kendall, a new measure of rank correlation, 1938, as chapter 0 section 0.9 states it, and the rank certificate’s statistic in readscope and openvector-bench.

Measurements

none

Failures and corrections

none

Invariance envelope

none declared

Machine checked

lean/DataMiningAsObservation/Kendall.lean, theorems not_both, card_add_le, tau_mem, concordant_comp, discordant_comp, tau_comp, discordant_self, at observation-data-mining f3914f0; what the check covers is stated in the book’s appendix C.

Used in

Data Mining as Observation primer S, chapters 0, 1, 8, 11.

Related

rank certificate; rank-faithful; Monotone Invariance Theorem; recall at k.

See also

Book equations stated beside the entry’s terms, not defining it: 11.2.

Ledger rows that cite the entry’s records without naming it: GO-B-Llama, GO-B-Llama-rematch.

Status

Generated 2026-09-10 by encyclopedia/generate.py; book at observation-data-mining f3914f0; the commit of every record is listed in the encyclopedia’s provenance.

← Jacobiankernel →