Skip to content

Design unified statistical correction and permutation inference - #562

Open
ljchang wants to merge 1 commit into
masterfrom
codex/unified-thresholding-design
Open

ljchang wants to merge 1 commit into
masterfrom
codex/unified-thresholding-design

Conversation

@ljchang

@ljchang ljchang commented Oct 15, 2026

Copy link
Copy Markdown
Member

Proposes a unified correction API for p-values and existing permutation nulls, plus a dedicated one-sample permutation-inference entry point and thin BrainData facade. Preserves the existing t-test dictionaries and threshold methods.

The design defines result records, family/tail semantics, cluster and TFCE scope, backend acceptance criteria, bounded-memory execution, compatibility, and tests. General inference stays in nltools; course simulations may live in dartbrains-tools. This PR contains a specification only and does not authorize or implement the new API.

Validation: uv run --no-sync poe ok passed (1,244 tests, 1 skipped); local Markdown links, code fences, placeholder scan, and git diff --check passed.

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant