Skip to content

Matplotlib Notes

Introduction to Matplotlib

  • Most popular data viz library
  • Created by John Hunter(Neurobiologist)
  • Initially created for EEG/ECoG Visualization Tool
  • Inspired from MATLAB

Matplotlib Architecture

+--------------------------+
| Scripting Layer(pyplot)  |
+--------------------------+
+--------------------------+
| Artist Layer(artist)     |
+--------------------------+
+--------------------------+
| Backend Layer(FigCanvas..|
+--------------------------+

Backend Layer(FigCanvas, renderer, events ...)

  1. FigureCanvas : matplotlib.backend_bases.FigureCanvas
    • Encompasses the area onto which the figure is drawn
  2. Renderer : matplotlib.backend_bases.Renderer
    • Knows how to draw on the FigureCanvas
  3. Event : matplotlib.backend_bases.Event
    • Handles user inputs such as keyboards strokes and mouse clicks

Artist Layer(artist)

  • Comprised of one main object - Artist
    • Know how to use the Renderer to draw on the Canvas
  • Responsible for : Title, lines, tock labals, and images, all correspond to individuals Artist instances
  • 2 types of artist objects :
      1. Primitive : Line2D, Rectangle, Circle, and Text
      1. Composite : Axis, Tick, Axes, and Figure
  • Each composite artist may contain other composite artists as well as primitive artists

Scripting Layer

  • Comprised mainly of pyplot, a scripting interface that is lighter that the Artist layer
  • Let's see how we can generate the same histogram of 10000 random values using the pyplot interface

Hands-On Matplotlib: Basic Plotting

  • Support by differents env : Python scripts, iPython shell, web app & servers ... jupyter nb
  • "Dynamic" ploting using BACKENDS : modify plot, costomize ...
  • use magic functions(%) to execute backend objects
    • magic function starts w/ % sign
  • Some backends :
    • %matplotlib inline : plot window within the browser and not in separeted window
    • %matplotlib notebook : allows to modify figure once is RENDERED !
  • Matplotlib - PANDAS
    • df.plot(kind="line")
    • df["x"].plot(kind="hist")

Dataset on Immigration to Canada

  • Dataset :

    • src : United Nations (45 countries)
    • annual data on the flows of international migrants
    • migrants to canada dataset

  • Import the dataset with pandas

  • to check the imported data : df.head, df.describe ...

Line Plots

  • continuos dataset
  • plot which displys information as series of data points called "markers" connected by the straight line segments
  • using pandas dataframe, map function to create an iterative point per axis value

References

  • https://matplotlib.org/
  • https://fr.wikipedia.org/wiki/Matplotlib
  • aosabook.org/en/matplotlib.html