Install

First, add the package:

import Pkg
Pkg.add("Plots")  # ≡ `PlotsBase` + `GR` backend

Pkg.add("PlotBase", "PythonPlot")  # `PlotsBase` + `PythonPlot` backend, avoids installing the `GR` backend

# if you want the latest features:
Pkg.pkg"add Plots#master"

The GR backend is included by default, but you can install additional plotting packages if you need a different backend.

Tier 1 support backends (in alphabetical order):

Pkg.add("GR")
# You do not need to add this package because it is the default backend and
# therefore it is automatically installed with Plots.jl. Note that you might
# need to install additional system packages if you are on Linux, see
# https://gr-framework.org/julia.html#installation

Pkg.add("UnicodePlots")  # simplest terminal based backend (guaranteed to work from a cluster, e.g. without X forwarding)

Pkg.add("PGFPlotsX")  # you need to have LaTeX installed on your system

Pkg.add("PlotlyJS"); Pkg.add("PlotlyBase")
# Note that you only need to add this if you need Electron windows and
# additional output formats, otherwise `plotly()` comes shipped with Plots.jl.
# In order to have a good experience with Jupyter, refer to Plotly-specific
# Jupyter installation (https://github.com/plotly/plotly.py#installation)

Pkg.add("PythonPlot")  # depends only on PythonPlot package

Pkg.add("Gaston")  # Gnuplot based backend

Learn more about backends here.

Finally, you may wish to add some extensions from the Plots ecosystem:

Pkg.add("StatsPlots")
Pkg.add("GraphRecipes")

Initialize

using Plots # or StatsPlots
# using GraphRecipes  # if you wish to use GraphRecipes package too

# or alternatively
import PythonPlot  # select installed backend (triggered by packages extensions: https://docs.julialang.org/en/v1/manual/code-loading/#man-extensions)
using PlotsBase

Optionally, choose a backend and/or override default settings at the same time:

gr(size = (300, 300), legend = false)  # provide optional defaults
unicodeplots()                         # plot in terminal
pgfplotsx()
plotly(ticks=:native)                  # plotlyjs for richer saving options
pythonplot()                           # backends are selected with lowercase names