Module inference 

Functional two-sample inference tests.

This module provides fdars’ first standalone functional-inference surface: two-sample functional hypothesis tests built by reusing existing permutation, Hotelling-T², and bootstrap-band machinery elsewhere in the crate. It is additive and non-breaking — no existing public signature is altered by this module.

R baselines

The functions here mirror the table-stakes two-sample tests from the R FDA ecosystem:

• t_perm_test / f_perm_test — fda::tperm.fd / fda::Fperm.fd (permutation two-sample mean tests).
• two_sample_mean_test — the FPC-basis Hotelling-T² mean-equality test in the spirit of fda.usc mean/covariance equality tests.
• mean_scb / scb_two_sample_test — SCBmeanfd-style simultaneous confidence bands for the mean and the mean difference.

Conventions

Permutation tests take an explicit deterministic seed (StdRng::seed_from_u64(seed)) and default to n_perm = 999 at the call site. All public functions return Result<_, FdarError> and validate their inputs at entry. Result structs derive Debug, Clone, PartialEq and are serde-gated.

Structs

TestResult

Result of a functional two-sample hypothesis test.

Constants

DEFAULT_N_PERM

Default number of permutations for the permutation tests.

Functions

f_perm_test

Functional two-sample permutation F-test (fda::Fperm.fd).

flm_f_test

Overall-significance F-test for a fitted functional linear model.

flm_gof_test

Residual-based lack-of-fit (goodness-of-fit) test for a fitted functional linear model.

mean_scb

Simultaneous confidence band for the mean function (Degras).

oneway_anova_vstat

Asymptotic one-way functional ANOVA test (V-statistic).

scb_two_sample_test

Two-sample mean-equality test via a simultaneous confidence band for the mean difference (SCBmeanfd-style).

t_perm_test

Functional two-sample permutation t-test (fda::tperm.fd).

two_sample_mean_test

Functional two-sample mean-equality test via Hotelling-T² on a shared FPC basis (fda.usc-style mean equality).