Incorrect nyquist plot

[Avitan01]

I'm using the control package with python 3.10 in Pycharm and Jupiter lab.
I have stumbled upon a possible issue while trying to create a Nyquist plot for a simple system

import control as ct

s = ct.tf('s'))
plant_1 = 1/(s*(s + 5)*(s + 7)))
ct.nyquist_plot(plant_1)

The result I'm getting is this:

While in MATLAB I'm getting different results

clear all;
close all;
clc;


s = tf('s');
plant_1 = 1/(s*(s + 5)*(s + 7));
nyquistplot(plant_1)

Is there a limitation that might be causing this issue or am I missing something?

[bnavigator]

Look at the y-axis scale. The nyquist plot from control cuts off at max_curve_magnitude and scales the "infinity" loop for display within the plot boundaries.
https://python-control.readthedocs.io/en/0.9.2/generated/control.nyquist_plot.html#control.nyquist_plot

[henklaak]

Behaviour that will 'surprise' many, I guess. The documentation is not explicit about it and the default value of 20.

How about not drawing the large magnitudes at all? Just clip the traces at the limit with a ≫to indicate direction?

[murrayrm]

If you clip the traces, then it is harder to see the number of net encirclements. Especially if there is more than one. For example

sys = ct.tf([1], [1, 3, 2]) * ct.tf([1], [1, 0, 1])
ct.nyquist_plot(sys)

which generates:

[murrayrm]

We should definitely make sure all of this is well-documented, though.

[henklaak]

I'll take a stab at this (the documenting part), expect PR.