Visualization

Summary in a discourse thread by Chris Rackauckas

• Plots.jl : powerful and convenient visualization with multiple backends. See also Plots.jl docs

• PythonPlot.jl : matplotlib in Julia. See also matplotlib docs

• Makie.jl : see the Makie.jl docs for examples.

PythonPlot.jl tips¶

“Clean” installation¶

If you don’t want CondaPkg.jl install a full conda environment in every project, you can prepare a common environment by micromamba.

1. Install micromamba, the minimal Python package manager.

2. Create a conda environment, for example, micromamba create -n juliapy jupyter matplotlib -y.

3. Set up the following environment variables

~/.profile

export CONDA_JL_HOME="${HOME}/micromamba/envs/juliapy"
export JULIA_PYTHONCALL_EXE="${CONDA_JL_HOME}/bin/python"
export JULIA_CONDAPKG_BACKEND="Null"

4. Rebuild Conda.jl or PythonCall.jl if necessary.

] build Conda
] build PythonCall

Save as TIFF¶

Exporting pyplot figures to TIFF images with a specific dpi and LZW compression for publishing.

using PythonCall
plt.savefig("fig1.tif", dpi=300, pil_kwargs=pydict(Dict("compression" => "tiff_lzw")))

Change default options¶

import PythonPlot as plt
plt.matplotlib.rcParams["font.size"] = 14