Moran's GOF-test statistic negative
We regularily get negative Moran GOF-test statistics, even when there are only all observed event times (no censored observations).
This is not good as we use a chi-square distribution to deduce P-value.
It look like some ties are enough to see this. Here is an example:
fm0 <- delay_model(x = c(8, 8, 8, 8, 9, 9, 9, 10, 10, 11, 12, 12, 12, 14, 14, 14, 16), method = "MPSE")
#tieInfo <- rlang::env_get(rlang::fn_env(fm0$objFun), "tieInfo")
test_GOF(delayFit = fm0, method = "moran")$statistic