Monte Carlo / Simulated Annealing for Ackleys function

Given Ackleys function in n dimensions and some arbitrary starting point, run a Monte Carlo / simulated annealing calculation to find an optimum.. or maybe not. That is the point. You can play with things like the temperature to see when the system converges and when it behaves badly.

Install and run

I do not have a mac for testing, so this is limited to linux and windows.

One can download the source, or perhaps there is a pre-built binary It can be built with or without graphics.

Prerequisites

• You must have go installed. I have only tried with go back to version 1.16. It may well build with earlier versions. It does expect support for go.mod.
• You need a C compiler and cgo for the graphics.
• If you disable graphics, you do not need C or cgo, but cross compiling (linux-> windows) works.

Build

Put in the instructions for downloading

Graphics yes / no

By default, it builds with the graphical interface. To turn off graphics, you need a build tag, no_gfx.

Binaries

Where are they ?

Documentation

The real documentation is in go doc format.

BUGS

A chart from go-chart in png format can be displayed by fyne. A chart in svg format leads to parse errors.

The graphical interface follows no known style convention and looks like it was designed by a three-year old.

Still have not implemented parallel runs, but given the emphasis on the quick graphics, this may not be appropriate.