CVXR: Power-class – R documentation

Become an expert in R — Interactive courses, Cheat Sheets, certificates and more!

Power-class

The Power class.

Description

This class represents the elementwise power function f(x) = x^p. If expr is a CVXR expression, then expr^p is equivalent to Power(expr, p).

Usage

Power(x, p, max_denom = 1024)

## S4 method for signature 'Power'
to_numeric(object, values)

## S4 method for signature 'Power'
sign_from_args(object)

## S4 method for signature 'Power'
is_atom_convex(object)

## S4 method for signature 'Power'
is_atom_concave(object)

## S4 method for signature 'Power'
is_atom_log_log_convex(object)

## S4 method for signature 'Power'
is_atom_log_log_concave(object)

## S4 method for signature 'Power'
is_constant(object)

## S4 method for signature 'Power'
is_incr(object, idx)

## S4 method for signature 'Power'
is_decr(object, idx)

## S4 method for signature 'Power'
is_quadratic(object)

## S4 method for signature 'Power'
is_qpwa(object)

## S4 method for signature 'Power'
.grad(object, values)

## S4 method for signature 'Power'
.domain(object)

## S4 method for signature 'Power'
get_data(object)

## S4 method for signature 'Power'
copy(object, args = NULL, id_objects = list())

## S4 method for signature 'Power'
name(x)

Arguments

`x`	The Expression to be raised to a power.
`p`	A numeric value indicating the scalar power.
`max_denom`	The maximum denominator considered in forming a rational approximation of `p`.
`object`	A Power object.
`values`	A list of numeric values for the arguments
`idx`	An index into the atom.
`args`	A list of arguments to reconstruct the atom. If args=NULL, use the current args of the atom
`id_objects`	Currently unused.

Details

For p = 0, f(x) = 1, constant, positive.

For p = 1, f(x) = x, affine, increasing, same sign as x.

For p = 2,4,8,..., f(x) = |x|^p, convex, signed monotonicity, positive.

For p < 0 and f(x) =

x^p for x > 0
+∞x ≤q 0

, this function is convex, decreasing, and positive.

For 0 < p < 1 and f(x) =

x^p for x ≥q 0
-∞x < 0

, this function is concave, increasing, and positive.

For p > 1, p \neq 2,4,8,… and f(x) =

x^p for x ≥q 0
+∞x < 0

, this function is convex, increasing, and positive.

Methods (by generic)

to_numeric: Throw an error if the power is negative and cannot be handled.
sign_from_args: The sign of the atom.
is_atom_convex: Is p ≤q 0 or p ≥q 1?
is_atom_concave: Is p ≥q 0 or p ≤q 1?
is_atom_log_log_convex: Is the atom log-log convex?
is_atom_log_log_concave: Is the atom log-log concave?
is_constant: A logical value indicating whether the atom is constant.
is_incr: A logical value indicating whether the atom is weakly increasing.
is_decr: A logical value indicating whether the atom is weakly decreasing.
is_quadratic: A logical value indicating whether the atom is quadratic.
is_qpwa: A logical value indicating whether the atom is quadratic of piecewise affine.
.grad: Gives the (sub/super)gradient of the atom w.r.t. each variable
.domain: Returns constraints describng the domain of the node
get_data: A list containing the output of pow_low, pow_mid, or pow_high depending on the input power.
copy: Returns a shallow copy of the power atom
name: Returns the expression in string form.

Slots

x: The Expression to be raised to a power.
p: A numeric value indicating the scalar power.
max_denom: The maximum denominator considered in forming a rational approximation of p.

CVXR

Disciplined Convex Optimization

v1.0-10

Apache License 2.0 | file LICENSE

Authors

Anqi Fu [aut, cre], Balasubramanian Narasimhan [aut], David W Kang [aut], Steven Diamond [aut], John Miller [aut], Stephen Boyd [ctb], Paul Kunsberg Rosenfield [ctb]

Initial release