R - Convert objects of class circular back to numeric-CodePudding

I am using the package circular in R to fit a von Mises distribution to a set of data:

Mises <- mle.vonmises(x = angle, mu = NULL, kappa = NULL, bias = FALSE, control.circular = list())

The results are from the class circular. Is there a way to extract both mu and kappa to have them as numerics? I need to do that many times (I guess using group_by()), so writing down manually the displayed values of mu and kappa is not an option.

Thanks for the help!

CodePudding user response：

The output of mle.vonmises is a list.

> str(Mises)
List of 8
 $ call     : language mle.vonmises(x = x)
 $ mu       : 'circular' num -0.0264  # it is a numeric value
  ..- attr(*, "circularp")=List of 6
  .. ..$ type    : chr "angles"
  .. ..$ units   : chr "radians"
  .. ..$ template: chr "none"
  .. ..$ modulo  : chr "asis"
  .. ..$ zero    : num 0
  .. ..$ rotation: chr "counter"
 $ kappa    : num 5.21
 $ se.mu    : num 0.0654
 $ se.kappa : num 0.973
 $ est.mu   : num 1
 $ est.kappa: num 1
 $ bias     : logi FALSE
 - attr(*, "class")= chr "mle.vonmises"

We can use standard extractors to extract the list elements ie. $ or [[. Regarding conversion to numeric, the mu is already numeric class.

> class(Mises$mu)
[1] "circular" "numeric"

It also include some attributes with an additional class circular built on top of numeric which can be removed by either wrapping with as.numeric or as.vector or set the attributes to NULL

as.numeric(Mises$mu)
#[1] -0.02639894
Mises$kappa
[1] 5.214254

data

library(circular)
x <- rvonmises(n=50, mu=circular(0), kappa=5)
Mises <- mle.vonmises(x)