round.dual: Rounding of Numbers
In ADtools: Automatic Differentiation Toolbox

Rounding of Numbers

1 2	## S3 method for class 'dual' round(x, digits = 0)

`x`	A "dual" object.
`digits`	An integer indicating the number of decimal places.

The function 'round' does not have a derivative over the real line. The derivative will be kept unchanged. The reason of not dropping it is that sometimes one may need to round a matrix to correct floating-point errors. This is often used when an apparent symmetric matrix does not pass the symmetry test.

ADtools documentation built on Nov. 9, 2020, 5:09 p.m.