jax.scipy.special.owens_t#
- jax.scipy.special.owens_t(h, a)[source]#
Owen’s T function.
JAX implementation of
scipy.special.owens_t.Computes Owen’s T function:
\[T(h, a) = \frac{1}{2\pi} \int_0^a \frac{\exp\!\left(-\tfrac{1}{2}h^2(1+x^2)\right)}{1+x^2} \, dx\]Computed via 13-point Gauss-type quadrature on the canonical integral form (Patefield & Tandy 2000 method T5). The full 18-region dispatch from Patefield & Tandy is intentionally avoided because XLA evaluates every branch of a
where/selectunconditionally, which turns per-region dispatch into added cost rather than savings.- Parameters:
h (ArrayLike) – array_like, real-valued.
a (ArrayLike) – array_like, real-valued.
- Returns:
Array of Owen’s T values with dtype matching the promoted inputs.
- Return type:
See also