Broom tidies a number of lists that are effectively S3
objects without a class attribute. For example, stats::optim()
,
svd() and akima::interp()
produce consistent output, but
because they do not have a class attribute, they cannot be handled by S3
dispatch.
These functions look at the elements of a list and determine if there is
an appropriate tidying method to apply to the list. Those tidiers are
themselves are implemented as functions of the form tidy_<function>
or glance_<function>
and are not exported (but they are documented!).
If no appropriate tidying method is found, throws an error.
tidy_optim(x, ...)
x | A list returned from |
---|---|
... | Additional arguments. Not used. Needed to match generic
signature only. Cautionary note: Misspelled arguments will be
absorbed in |
This function assumes that the provided objective function is a negative log-likelihood function. Results will be invalid if an incorrect function is supplied.
tidy(o) glance(o)
Other list tidiers:
glance_optim()
,
list_tidiers
,
tidy_irlba()
,
tidy_svd()
,
tidy_xyz()
A tibble::tibble()
with columns:
The parameter being modeled.
The standard error of the regression term.
The value/estimate of the component. Results from data reshaping.
f <- function(x) (x[1] - 2)^2 + (x[2] - 3)^2 + (x[3] - 8)^2 o <- optim(c(1, 1, 1), f)