5.1.4.1. numdifftools.limits.CStepGenerator

class CStepGenerator(base_step=None, step_ratio=4.0, num_steps=None, step_nom=None, offset=0, scale=1.2, **options)

Generates a sequence of steps

where

steps = base_step * step_nom * (exp(1j*dtheta) * step_ratio) ** (i + offset)

for i = 0, 1, …, num_steps-1

Parameters

base_stepfloat, array-like, default None

Defines the minimum step, if None, the value is set to EPS**(1/scale)

step_ratioreal scalar, optional, default 4.0

Ratio between sequential steps generated.

num_stepsscalar integer, optional,

defines number of steps generated. If None the value is 2 * int(round(16.0/log(abs(step_ratio)))) + 1

step_nomdefault maximum(log(exp(1)+|x|), 1)

Nominal step where x is supplied at runtime through the __call__ method.

offsetreal scalar, optional, default 0

offset to the base step

use_exact_stepsboolean, default True.

If true make sure exact steps are generated.

scalereal scalar, default 1.2

scale used in base step.

path‘radial’ or ‘spiral’

Specifies the type of path to take the limit along. Default ‘radial’.

dtheta: real scalar, default pi/8

If the path is ‘spiral’ it will follow an exponential spiral into the limit, with angular steps at dtheta radians.

__init__(base_step=None, step_ratio=4.0, num_steps=None, step_nom=None, offset=0, scale=1.2, **options)

Methods

__init__([base_step, step_ratio, num_steps, ...])
step_generator_function(x[, method, n, order])	Step generator function

Attributes

base_step	Base step defines the minimum or maximum step when offset==0.
dtheta	Angular steps in radians used for the exponential spiral path.
min_num_steps	Minimum number of steps required given the differentiation method and order.
num_steps	The number of steps generated
scale	Scale used in base step.
step_nom	Nominal step
step_ratio	Ratio between sequential steps generated.