Summarizes the intercept, slope, and residual SD of a calibration. It works the same way for bootstrap calibrations (cal.ols() and others) and Bayesian calibrations (cal.bayesian() and cal.ion.bayesian()).
Arguments
- fit
A
data.frameof bootstrap replicates or posterior draws, a result fromcal.bayesian()orcal.ion.bayesian(), or a singlestanfitfrom it.- model
The model to use when
fitholds more than one, such as"errors".- level
Probability covered by the interval.
Value
A data.frame with one row per parameter and the columns parameter, estimate (mean), SE, median, lower, and upper. Mixed models also have a Material column.
Details
SE is the standard error of each estimate. It is the SD of the bootstrap replicates or of the posterior draws. sigma is itself an SD: the scatter of individual samples around the calibration line. Its SE describes how well that scatter is known.
Examples
ds <- cal.dataset(error = "S1", nobs = 50, seed = 1)
ols <- cal.ols(ds, replicates = 200)
cal.summary(ols)
#> parameter estimate SE median lower upper
#> 1 alpha 0.26952294 0.0112427446 0.26848511 0.249811220 0.29615864
#> 2 beta 0.03664000 0.0008750491 0.03672634 0.034499765 0.03827147
#> 3 sigma 0.01202395 0.0013887962 0.01205333 0.009551602 0.01495939
