scikit-tda

scikit-tda implements nothing. It exists so that one install brings the interoperating Python TDA components together, which makes it the fixture to reach for when the question is whether the ecosystem coheres rather than whether one engine is fast. See scikit-tda-environment.

Its run closure is on conda channels with one exception, tadasets, which tadasets-recipe supplies and has a staged-recipes pull request open for. Once that lands, the only thing still holding scikit-tda-environment at L1 is this recipe.

This one has never been built. Nothing here records a rattler-build run, and the environment has no solved lock either, so the recipe is a generated starting point — modelled on rattler-build generate-recipe pypi scikit-tda — rather than a verified one. It is pure Python and MIT, so there is no reason to expect trouble; there is also no evidence there is none. Building it is the cheap next step, and it should happen before the tadasets PR makes this the last blocker.