README.md

go-gnuplot

Simple-minded functions to work with gnuplot. go-gnuplot runs gnuplot as a subprocess and pushes commands via the STDIN of that subprocess.

See http://www.gnuplot.info for more informations on the exact semantics of these commands.

This is a fork of sbinet/go-gnuplot. The fork is motivated by the lack of maintenance to the original repo and the need for 3D and quiver plotting in scientific computing not yet provided by native Golang plotting libraries. This version will aim to extend upon and modernize the original implementation without changing the base API so that this can remain a drop-in replacement. See the issues for specific planned changes or if you want to see a change file an issue against the repo.

Installation

The go-gnuplot package is go get installable:

$ go get github.com/ckitagawa/go-gnuplot

Example

package main

import (
  "github.com/ckitagawa/go-gnuplot"
  "fmt"
)

func main() {
	fname := ""
	persist := false
	debug := true

	p,err := gnuplot.NewPlotter(fname, persist, debug)
	if err != nil {
		err_string := fmt.Sprintf("** err: %v\n", err)
		panic(err_string)
	}
	defer p.Close()

	p.PlotX([]float64{0,1,2,3,4,5,6,7,8,9,10}, "some data")
	p.CheckedCmd("set terminal pdf")
	p.CheckedCmd("set output 'plot002.pdf'")
	p.CheckedCmd("replot")

	p.CheckedCmd("q")
	return
}

Motivation

Gonum now makes it feasible to author scientific and engineering computing code in Go. This is great news because it enables more efficient, compiled and type-safe code that is not nearly as easy to achieve in Python. Furthermore, it is less of a headache to get a Go program running than C or C++ and is usually more portable. Moreover, it has great concurrency primitives which are a real boon to doing numeric computing more efficiently.

That being said, this project is slightly counter-intuitive in the sense it limits portability to platforms supporting gnuplot and currently doesn't play nice with goroutines. Additionally, native graphing packages like wcharczuk/go-chart and gonum/plot already exist; however, none of them have mature 3D plotting capabilities nor strong support for quivers and streamlines. Unfortunately, these are features required in engineering and scientific computing on a fairly regular basis. As a result, this package needs to live on a bit longer to service those in need of these specialized charts. That said, it is my sincere hope that this library will not be necessary for much longer.

Plans

• Add unit tests, a challenge which will probably require mocking out gnuplot
• Extend the original API to allow saving to files without Cmd or CheckedCmd. Also create methods for common actions such as replot.
• Support gnuplot > 5 as the original implementation was written for gnuplot 4