
delta response (wave-3 mean):
  delta=-0.357 (x0.70) -> 6.2755
  delta=-0.163 (x0.85) -> 7.1240
  delta=+0.000 (x1.00) -> 7.9230
  delta=+0.140 (x1.15) -> 8.4145
  delta=+0.336 (x1.40) -> 9.6555
.
negbin omitted-interaction (expected failure): reps=500 fail=0 truth=+0.4900 bias=-0.2392 (mc se 0.0072, std -1.483) rmse=0.2885 sd=0.1613 se_ratio=1.631 coverage=0.950 width=1.0549 fmi=0.345
.
negbin MNAR-under-MAR (expected failure): reps=500 fail=10 truth=+7.4109 bias=-3.8145 (mc se 0.0233, std -7.381) rmse=3.8493 sd=0.5168 se_ratio=1.158 coverage=0.018 width=2.6858 fmi=0.723
.
gaussian omitted-auxiliary (expected failure): reps=500 fail=0 truth=+3.0000 bias=-0.1575 (mc se 0.0064, std -1.100) rmse=0.2128 sd=0.1432 se_ratio=0.995 coverage=0.798 width=0.5835 fmi=0.419
.
gaussian mcar: reps=500 fail=0 truth=+1.2000 bias=-0.0136 (mc se 0.0090, std -0.068) rmse=0.2016 sd=0.2013 se_ratio=0.983 coverage=0.948 width=0.7885 fmi=0.243
.
gaussian mar: reps=500 fail=0 truth=+1.2000 bias=+0.0029 (mc se 0.0099, std +0.013) rmse=0.2221 sd=0.2223 se_ratio=0.958 coverage=0.944 width=0.8569 fmi=0.345
.
gaussian intermittent mar: reps=500 fail=0 truth=+1.2000 bias=-0.0154 (mc se 0.0084, std -0.081) rmse=0.1893 sd=0.1888 se_ratio=1.029 coverage=0.964 width=0.7724 fmi=0.216
.
negbin mar: reps=500 fail=0 truth=+0.3986 bias=+0.0013 (mc se 0.0116, std +0.005) rmse=0.2594 sd=0.2597 se_ratio=1.030 coverage=0.960 width=1.0786 fmi=0.387
.
negbin intermittent mar: reps=500 fail=0 truth=+0.3986 bias=-0.0183 (mc se 0.0109, std -0.075) rmse=0.2437 sd=0.2432 se_ratio=1.003 coverage=0.948 width=0.9692 fmi=0.254
.
negbin high-missingness: reps=500 fail=4 truth=+0.3986 bias=+0.0391 (mc se 0.0166, std +0.106) rmse=0.3709 sd=0.3692 se_ratio=0.937 coverage=0.942 width=1.4565 fmi=0.627
.
=============================== warnings summary ===============================
tests/simulation/test_expected_failures.py::TestMnarIsNotRescuedByMarImputation::test_wave3_mean_remains_biased
tests/simulation/test_negbin_coverage.py::TestNegBinImputerFrequentistValidity::test_high_missingness_stress
tests/simulation/test_negbin_coverage.py::TestNegBinImputerFrequentistValidity::test_high_missingness_stress
tests/simulation/test_negbin_coverage.py::TestNegBinImputerFrequentistValidity::test_high_missingness_stress
  /Users/aarondanielson/Dropbox/longmi/src/longmi/impute/negbin.py:420: RuntimeWarning: overflow encountered in exp
    tau = float(np.exp(theta[p + 1]))

tests/simulation/test_expected_failures.py::TestMnarIsNotRescuedByMarImputation::test_wave3_mean_remains_biased
  /Users/aarondanielson/Dropbox/longmi/src/longmi/impute/negbin.py:430: RuntimeWarning: overflow encountered in exp
    lam = rng.gamma(shape=kappa, scale=np.exp(log_mu) / kappa)

-- Docs: https://docs.pytest.org/en/stable/how-to/capture-warnings.html
10 passed, 5 warnings in 2965.58s (0:49:25)
