OUT OF DATE Use Plots.jl and gr() for now on a generic setup.
At this point, plotting in Julia is a mess. Lots of great work is being done, but it will take time.
A key principle is that you do not need to use the same plotting library for interactive graphs in Jupyter/etc. while creating your algorithm and for creating publication quality plots.
- PyPlot.jl: A wrapper for matplotlib library. It good for Jupyter as well as exporting figures for print, and is fast to use. The documentation is poor, as users generally read the Python documents directly to figure out features.
- PlotlyJS.jl: A wrapper for the interactive plotly.js.js library. It is perfect for Jupyter, very fast, interactive, and has decent documentation. It isn't really intended for print.
- Plots.jl: The goal of this was to create a common front-end interface for plotting, where you can switch out the various backends which actually generate the plots (i.e., can choose PlotlyJS, PyPlot, and others). It is a great idea which could conceivably work as the baseline plotting library, but it is incredibly slow on first use. It tends to be feature rich and well-documented.
At this point, consider sticking with PyPlot.jl as a starting point, and PlotlyJS.jl as you get comfortable in Jupyter and with a desktop setup
- PyPlot.jl: The baseline documentation emphasizses how to translate the Python docs to PyPlot, so learn through examples instead
- Tutorial in QuantEcon
- A large number of examples are in https://gist.github.com/gizmaa/7214002
- See the tutorial in PaulSoderlind/JuliaTutorial/PyPlot
- Some examples of animations
- genkuroki animations
- genkuroki animations
- 2D three-body problem -
- 3D n-body problem- potentially out of date
- PlotlyJS.jl The baseline documentation is good
- Tutorial in QuantEcon
- Best learned through the examples in the documentation
- PlotlyJS on your desktop will take a little work. See the documentation.
- Plots.jl:
- Tutorial in QuantEcon
- The baseline documentation is excellent
- PaulSoderlind/JuliaTutorial/Plots