## kernels.leave_one_out_coverage()


Coverage of intervals built for each calibration forecast from the OTHER


Usage

``` python
kernels.leave_one_out_coverage(
    calibration,
    *,
    levels=(0.8, 0.95),
)
```


units of its stratum (the R study's cross-company historical coverage).

The quantiles are numpy's `median_unbiased` method, which is R's type 8.
