EditURL = "https://github.com/TRAVIS_REPO_SLUG/blob/master/"
+++
title = "Get started"
description = "Up and running in under five minutes"
weight = -1000
draft = false
toc = true
bref = "For this guide, we're going to walk through a simple example."
+++
This guide assumes that you have Julia 1.0 installed and running.
If this is not the case follow the instructions at julialang.org.
HomotopyContinuation.jl is available through the Julia package manager by
pkg> add HomotopyContinuation
This can be accomplished as follows
using HomotopyContinuation
@polyvar x y; # declare the variables x and y
solve([x^2+2y, y^2-2], report_progress=false) #hide
result = solve([x^2+2y, y^2-2])
Let us see what is the information that we get. Four paths were attempted to be solved, four of which were completed successfully. Since we tried to solve an affine system, the algorithm checks whether there are solutions at infinity: in this case there are none. With solutions at infinity we mean solutions of the homogenization of the system which are no solutions of the affine system. None of the solutions are singular and two of them are real. To access the first solution in the array we write
# Assume we are only interested in the *real* solutions. We can obtain these by
real(result)
where we can see that there are 2 real solutions,
real(result)[1]
The returncode tells us that the pathtracking was successfull. What do the other entries of that table tell us? Let us consider the most relevant:
-
solution
: the zero that is computed (here it is $(2^{\frac34},-\sqrt{2})$). -
singular
: boolean that shows whether the zero is singular. -
residual
: the computed value of|f([1.68179+1.27738e-17im, -1.41421-1.18454e-17im])|
. -
windingnumber
: This is a lower bound on the multiplicity of the solution. A windingnumber greater than 1 indicates that the solution is singular.
To extract the solutions you can do
results(solution, result)
or
results(solution, result, onlyreal=true)
to obtain only real solutions.
This page was generated using Literate.jl.