Expand description
Group statistics on feature tables: GLM t with Freedman–Lane permutation, Welch’s t, sign flipping for paired designs, FWER by the maximum statistic, FDR, and the network-based statistic (docs/statistics.md).
Structs§
- Component
- A connected set of suprathreshold edges within one network.
- Edge
- Where an edge feature sits: which network (component k of the connectivity) and which nodes.
- Glm
- A design matrix (observations × regressors) and the regressor being tested.
- Permutation
Test - Test
Result - Per-feature results.
Enums§
- Exchange
- What the observations may be rearranged by under the null hypothesis.
- Fdr
- Which false discovery rate procedure.
- Statistic
- The statistic tested.
- Tail
- Which differences count as evidence.
Functions§
- adjust
- Adjusted p-values (q-values) in the original order. NaN p-values stay NaN and are not counted.
- incomplete_
beta - Regularised incomplete beta I_x(a, b) by its continued fraction (Lentz’s method), using the symmetry I_x(a, b) = 1 − I_{1−x}(b, a) where the fraction converges faster.
- t_upper
- Upper tail probability P(T > t) for Student’s t with
dfdegrees of freedom. - welch_
df - Welch–Satterthwaite degrees of freedom per feature.
- welch_t
- Welch’s t of group
aminus the rest, for each feature (column ofy).