## kernels.calibrate()


Pool the named horizons' errors by unit size and centre each pool.


Usage

``` python
kernels.calibrate(
    residuals,
    *,
    horizons,
    n_strata=4,
    min_per_stratum=40,
)
```


A stratum holding fewer than `min_per_stratum` residuals is refused by name: resampling from a handful of errors would report a spread that is mostly an accident of which units fell into that stratum.
