Reports what a self-test is run to find out: how many simulations the estimator brought back to an optimum, and under what the replicates were generated. The returned list cannot say the first on its own: non-converged runs are dropped before it is returned, so the number of fits returned is not the number attempted.
Usage
# S3 method for class 'Rceattle_selftest'
print(x, rate = 0.9, ...)Arguments
- x
A
"Rceattle_selftest"object fromself_test().- rate
Fraction of simulations that must converge before the run counts as clean. Default
0.9, matchingjitter()'s.- ...
Currently unused.
Details
The status is the convergence RATE, since that is the quantity the run
produces: FAIL if nothing converged, WARN below rate, NOTE if any
replicate was dropped, OK otherwise. The table beneath tallies each
returned fit's own $convergence$status, which is a separate question,
a replicate can reach a zero gradient and still hold a NOTE or WARN,
and those are not folded into the header for that reason.
The last line says what generated the replicates, because it decides what the
spread means. With processes held fixed (the default) sim_mod() redraws the
observations alone, so the spread reflects observation error only and is a
lower bound on estimation uncertainty, do not read it against the model's
own uncertainty bands, which include process error. See
Interpreting the spread in self_test().