Closed tomwhite closed 2 weeks ago
Looks like it may be due to stricter checks in NumPy 2.0.0.
tau_p = np.uint8(0), lambda_p = np.float64(0.0), ploidy_p = np.uint64(2) kinship_pp = 0.0, tau_q = np.uint8(2), lambda_q = np.float64(0.167) ploidy_q = np.uint64(4), kinship_qq = np.float64(0.307798800184587) kinship_pq = 0.0 @numba_jit def _hamilton_kerr_inbreeding_non_founder( tau_p: int, lambda_p: float, ploidy_p: int, kinship_pp: float, tau_q: int, lambda_q: float, ploidy_q: int, kinship_qq: float, kinship_pq: float, ) -> float: # pragma: no cover """Calculate the inbreeding coefficient of a non founder individual i with parents p and q. """ pat = ( tau_p > * (tau_p - 1) * (lambda_p + (1 - lambda_p) * ((ploidy_p * kinship_pp - 1) / (ploidy_p - 1))) ) E RuntimeWarning: overflow encountered in scalar subtract
From https://github.com/sgkit-dev/sgkit/actions/runs/9540232291/job/26291678469
Looks like it may be due to stricter checks in NumPy 2.0.0.
From https://github.com/sgkit-dev/sgkit/actions/runs/9540232291/job/26291678469