DEV Community

kojix2
kojix2

Posted on • Edited on

Introduction to GR.rb - data visualization with Ruby

GitHub logo kojix2 / GR.rb

Ruby wrapper for the GR framework

GR.rb

Gem Version CI Gitter Chat Docs Latest

rdatasets-1 stem histogram barplot scatter3 volume griddata 2darray 2dpolararray hexbin rdatasets-2 rdatasets-3 surface face shade

πŸ“Š GR framework - powerful visualization library - for Ruby

Installation

First, install GR. Then install gr-plot gem.

gem install gr-plot
Enter fullscreen mode Exit fullscreen mode

pkg-config will detect the location of the shared library. Otherwise, you need to specify the environment variable.

export GRDIR="/your/path/to/gr"
Enter fullscreen mode Exit fullscreen mode

Quick Start

πŸ‘‰ Wiki -plotting functions

require 'gr/plot'

x = [0, 0.2, 0.4, 0.6, 0.8, 1.0]
y = [0.3, 0.5, 0.4, 0.2, 0.6, 0.7]

# show the figure
GR.plot(x, y)

# Save the figure in PNG format.
GR.savefig("figure.png")
Enter fullscreen mode Exit fullscreen mode

GR.rb supports Jupyter Notebook / Lab.

API Overview

GR::Plot - A simple, matlab-style API

require 'gr/plot'
GR.plot(x, y)
Enter fullscreen mode Exit fullscreen mode

List of available functions.

plot step plot3 polar scatter scatter3 stem barplot histogram polarhistogram hexbin contour contourf…

GR.rb is a Ruby visualization tool I've been working since last fall. This is the binding of GR Framework, which is the default backend for Plots.jl. Here you can see a gallery of some of our examples.

Getting ready

Install GR Framework.
Set environment variable GRDIR.

export GRDIR="/your/path/to/gr"
gem install ruby-gr
```



you'll need the `ruby-gr`, `numo-narray` gems.



```
gem install ruby-gr
```





```ruby
require 'gr/plot'
require 'numo/narray'
```



## Line Plots



```ruby
x = Numo::DFloat.linspace(0, 10, 1001)
y = Numo::NMath.sin(x)

GR.plot(x, y)
```



![](https://user-images.githubusercontent.com/5798442/80851620-c1d48900-8c5d-11ea-8c78-376bb3b2d644.png)



```ruby
x = Numo::DFloat.linspace(0, 10, 51)
y = Numo::NMath.sin(x)

GR.step(x, y)
```



![](https://user-images.githubusercontent.com/5798442/80851791-c188bd80-8c5e-11ea-865f-527db1846a61.png)



```ruby
x = Numo::DFloat.linspace(0, 30, 1000)
y = Numo::NMath.cos(x) * x
z = Numo::NMath.sin(x) * x

GR.plot3(x, y, z)
```



![](https://user-images.githubusercontent.com/5798442/80851853-188e9280-8c5f-11ea-9416-44a558706607.png)



```ruby
angles = Numo::DFloat.linspace(0, 2 * Math::PI, 40)
radii = Numo::DFloat.linspace(0, 20, 40)

GR.polar(angles, radii)
```



![](https://user-images.githubusercontent.com/5798442/80851874-47a50400-8c5f-11ea-89ff-1186d5942801.png)

## Scatter Plots



```ruby
x = Numo::DFloat.linspace(0, 1, 51)
y = x - x**2
s = x * 200

GR.scatter(x, y, s)
```



![](https://user-images.githubusercontent.com/5798442/80852173-73c18480-8c61-11ea-9625-cdaf4c07a659.png)



```ruby
x = Numo::DFloat.linspace(0, 1, 51)
y = x - x**2
s = Numo::DFloat.linspace(50, 300, x.size)
c = Numo::DFloat.linspace(0, 255, x.size)

GR.scatter(x, y, s, c)
```



![](https://user-images.githubusercontent.com/5798442/80852185-850a9100-8c61-11ea-8e41-3a49cae38b9d.png)



```ruby
x = 2 * Numo::DFloat.new(100).rand - 1
y = 2 * Numo::DFloat.new(100).rand - 1
z = 2 * Numo::DFloat.new(100).rand - 1
c = 999 * Numo::DFloat.new(100).rand + 1

GR.scatter3(x, y, z, c)
```



![](https://user-images.githubusercontent.com/5798442/80852233-e3d00a80-8c61-11ea-805a-2b1356425dba.png)

## Stem Plots



```ruby
x = Numo::DFloat.linspace(-2, 2, 40)
y = x ** 3 + x ** 2 + x + 6

GR.stem(x, y)
```



![](https://user-images.githubusercontent.com/5798442/80852331-b0da4680-8c62-11ea-99ee-e393acf86028.png)

## Bar Plots



```ruby
continents = ["Africa", "America", "Asia", "Europe", "Oceania"]
population_2010 = [1044, 944, 4170, 735, 36]

GR.barplot(continents, population_2010)
```



![](https://user-images.githubusercontent.com/5798442/80907904-b49ac580-8d55-11ea-87f4-59dc14196c26.png)

## Histograms



```ruby
x = Numo::DFloat.new(10_000).rand_norm

GR.histogram(x)
```



![image](https://user-images.githubusercontent.com/5798442/80852386-1f1f0900-8c63-11ea-91e6-a01d9c2530cb.png)



```ruby
x = Numo::DFloat.new(10_000).rand_norm
y = Numo::DFloat.new(10_000).rand_norm

GR.hexbin(x, y)
```



![image](https://user-images.githubusercontent.com/5798442/80852515-2a266900-8c64-11ea-8eec-84fa20379174.png)

## Contour Plots



```ruby
x = 8 * Numo::DFloat.new(60).rand - 4
y = 8 * Numo::DFloat.new(60).rand - 4
z = Numo::NMath.sin(x) + Numo::NMath.cos(y)

GR.contour(x, y, z)
```



![image](https://user-images.githubusercontent.com/5798442/80852802-635fd880-8c66-11ea-90e8-bd8f6489d3ca.png)



```ruby
_x = Numo::DFloat.linspace(-2, 2, 40)
_y = Numo::DFloat.linspace(0, Math::PI, 20)
x = (_x.expand_dims(0) * Numo::DFloat.ones(_y.size, 1)).flatten
y = (_y.expand_dims(1) * Numo::DFloat.ones(1, _x.size)).flatten
z = (Numo::NMath.sin(x) + Numo::NMath.cos(y)).flatten

GR.contour(x, y, z)
```



![](https://user-images.githubusercontent.com/5798442/80852673-50003d80-8c65-11ea-95bf-ddd3d363692a.png)



```ruby
x = 8 * Numo::DFloat.new(100).rand - 4
y = 8 * Numo::DFloat.new(100).rand - 4
z = Numo::NMath.sin(x) + Numo::NMath.cos(y)

GR.contourf(x, y, z)
```



![](https://user-images.githubusercontent.com/5798442/80852853-eb45e280-8c66-11ea-857b-4abfd5675374.png)



```ruby
x = 8 * Numo::DFloat.new(300).rand - 4
y = 8 * Numo::DFloat.new(300).rand - 4
z = Numo::NMath.sin(x) + Numo::NMath.cos(y)

GR.tricont(x, y, z)
```



![](https://user-images.githubusercontent.com/5798442/80852926-96569c00-8c67-11ea-889d-852b585ceafc.png)

## Surface Plots



```ruby
x = 8 * Numo::DFloat.new(100).rand - 4
y = 8 * Numo::DFloat.new(100).rand - 4
z = Numo::NMath.sin(x) + Numo::NMath.cos(y)

GR.surface(x, y, z)
```



![](https://user-images.githubusercontent.com/5798442/80854132-9dce7300-8c70-11ea-9ae6-3d4a7a0d2f88.png)



```ruby
x = 8 * Numo::DFloat.new(100).rand - 4
y = 8 * Numo::DFloat.new(100).rand - 4
z = Numo::NMath.sin(x) + Numo::NMath.cos(y)

GR.trisurf(x, y, z)
```



![](https://user-images.githubusercontent.com/5798442/80854210-264d1380-8c71-11ea-9d9e-7e8182f06c1b.png)



```ruby
x = 8 * Numo::DFloat.new(1000).rand - 4
y = 8 * Numo::DFloat.new(1000).rand - 4
z = Numo::NMath.sin(x) + Numo::NMath.cos(y)

GR.wireframe(x, y, z)
```



![](https://user-images.githubusercontent.com/5798442/80854250-75934400-8c71-11ea-9278-352c74d97b82.png)

## Volume Rendering



```ruby
z = Numo::DFloat.new(50, 50, 50).rand_norm

GR.volume(z)
```



![](https://user-images.githubusercontent.com/5798442/80854306-ffdba800-8c71-11ea-984e-192533d0a2ea.png)

## Heatmaps



```ruby
x = Numo::DFloat.linspace(-2, 2, 40)
y = Numo::DFloat.linspace(0, Math::PI, 20)
z = Numo::NMath.sin(x).expand_dims(0).transpose + Numo::NMath.cos(y)

GR.heatmap(z)
```



![image](https://user-images.githubusercontent.com/5798442/80854477-a5434b80-8c73-11ea-8a77-8f6f23085781.png)



```ruby
n = 1_000_000
x = Numo::DFloat.new(n).rand_norm
y = Numo::DFloat.new(n).rand_norm

GR.shade(x, y)
```



![](https://user-images.githubusercontent.com/5798442/80854645-c48ea880-8c74-11ea-8364-7351e11b60c7.png)

## Images



```ruby
x = Numo::DFloat.linspace(-1, 1, 20)
y = Numo::DFloat.linspace(0, Math::PI, 30)
z = Numo::NMath.sin(x).expand_dims(0).transpose + Numo::NMath.cos(y)

GR.imshow(z)
```



![](https://user-images.githubusercontent.com/5798442/80857969-2bba5600-8c91-11ea-850c-e682ebf1f843.png)

## Isosurfaces



```ruby
s = Numo::DFloat.linspace(-0.6, 0.6, 50)
v = 1 - ((s**2 +  (s**2).expand_dims(0).transpose).expand_dims(2) + s.reshape(1, 1, true)**2)**0.5

GR.isosurface(v)
```



![](https://user-images.githubusercontent.com/5798442/80867161-94bfbf00-8ccd-11ea-9405-9293634ae030.png)

## Subplots



```ruby
require 'gr/plot'

x = [1,2,3,4,5,6,7,8,9,10]
y = x.shuffle

GR.barplot x, y, GR.subplot(2, 2, 1)
GR.stem    x, y, GR.subplot(2, 2, 2)
GR.step    x, y, GR.subplot(2, 2, 3)
GR.plot    x, y, GR.subplot(2, 2, 4)
```



![](https://user-images.githubusercontent.com/5798442/80975253-cfa42d00-8e5c-11ea-8f75-4644a861002a.png)

## Resources
* https://github.com/red-data-tools/GR.rb/wiki/
Enter fullscreen mode Exit fullscreen mode

Top comments (0)