load_integration_benchmarks

datasets.load_integration_benchmarks()

Load the harmonized scIB, OpenProblems and Tran integration benchmark set.

Reads the bundled scIB raw scores and Tran ranks, plus the OpenProblems batch_integration scores already shipped, restricts to the five common methods and the four shared metrics (ARI, ASW, kBET, LISI), and reduces each benchmark-dataset-metric cell to a rank within the common methods. The result feeds the cross-benchmark analysis.

Returns

Type	Description
IntegrationBenchmarks