tree 17dfcc60a297c1365def3958396745dd521caa78
parent eed0858ff96597b843de708c012a42ec12af790b
author Marc Kupietz <kupietz@ids-mannheim.de> 1788237187 +0200
committer Marc Kupietz <kupietz@ids-mannheim.de> 1788237187 +0200

Readme: document comparing collocates across virtual corpora

Describe the multi-VC comparison that collocationAnalysis() performs
when vc is given as a named vector: which comparison columns are added
per association score, and the two properties that are easy to trip
over when using them, namely that the result holds one row per
collocate *and* virtual corpus with the comparison columns repeated on
each, and that scores of collocates attested in only one of the
compared corpora are imputed from a floor value.

The example was run against the live DeReKo instance and the shown
dplyr pipeline was verified on its result. No result table is included:
without an access token the comparison only sees the freely available
subset, where the compared corpora are covered too unevenly for the
numbers to illustrate anything.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Change-Id: If845cf72c3de90d13a0bab68fa6e8928c36330cd