Examples

Basic plotting

Individual variables from InferenceData and ArviZ.Dataset can be plotted using any series type that can operate on the underlying draws.

Plot histogram (Note that the histogram recipe is defined generally in Plots.jl).

using ArviZ, ArviZPlots, Plots

data = load_arviz_data("centered_eight")
histogram(data; legend=false, bins=20, titlefontsize=8)

Plot estimated density of posterior and then prior.

distplot(data; fill=true, alpha=0.5, var_names=["mu"], label="posterior")
distplot!(data; fill=true, alpha=0.5, var_names=["mu"], groupname=:prior, label="prior")

Plot using the custom :arviz-darkgrid theme (see Themes).

theme(:arviz_darkgrid)
distplot(data; fill=true, var_names=["mu"])

Dist Plot

Plot an integer distribution

using ArviZPlots, Distributions, Plots
theme(:arviz_darkgrid)

a = rand(Poisson(4), 1000)
distplot(a)

Plot a continuous distribution

b = randn(1000)
distplot(b)

Add a rug under the Gaussian distribution

distplot(b; showrug=true)

Plot as the cumulative distribution

distplot(b; showrug=true, cumulative=true)

Compare with the empirical cumulative distribution function

c = randn(100)
distplot(c; cumulative=true, label="KDE CDF")
distplot!(c; kind=:ecdf, label="ECDF", legend=:topleft)

See distplot

Energy Plot

Plot a default energy plot

using ArviZ, ArviZPlots, Plots
theme(:arviz_darkgrid)

data = load_arviz_data("centered_eight")
energyplot(data)

Represent energy plot via histograms

energyplot(data; kind=:hist, strokewidth=0)

See energyplot

KDE Plot

Plot default KDE

using ArviZ, ArviZPlots, Plots
theme(:arviz_darkgrid)

non_centered = load_arviz_data("non_centered_eight")
mu_posterior = vec(non_centered.posterior["mu"].values)
tau_posterior = vec(non_centered.posterior["tau"].values)
kdeplot(mu_posterior)

Plot KDE with rugplot

kdeplot(mu_posterior; showrug=true)

Plot the estimated cumulative distribution

kdeplot(mu_posterior; cumulative=true)

Rotate plot 90 degrees and fill

kdeplot(mu_posterior; rotated=true, fill=true)

See kdeplot

2d KDE Plot

Plot 2d contour KDE

using ArviZ, ArviZPlots, Plots
theme(:arviz_darkgrid)

non_centered = load_arviz_data("non_centered_eight")
mu_posterior = vec(non_centered.posterior["mu"].values)
tau_posterior = vec(non_centered.posterior["tau"].values)
kde2dplot(mu_posterior, tau_posterior)

Plot 2d contour KDE, with filling.

kde2dplot(mu_posterior, tau_posterior; fill=true)

Plot 2d contour KDE, set the number of levels to 3.

kde2dplot(mu_posterior, tau_posterior; levels=3)

Plot 2d smooth KDE

kde2dplot(mu_posterior, tau_posterior; contour=false)

See kde2dplot

Pair Plot

KDE Pair Plot

using ArviZ, ArviZPlots, Plots

centered = load_arviz_data("centered_eight")
coords = Dict("school"=> ["Choate", "Deerfield"])
pairplot(
    centered;
    var_names=["theta", "mu", "tau"],
    kind=:kde,
    coords=coords,
    showdivergences=true,
)

Histogram pair plot

pairplot(centered; var_names=["theta", "mu"], kind=:hist, coords=coords)

Pair plot showing divergences and select variables with regular expressions

pairplot(
    centered;
    var_names=["^t", "mu"],
    filter_vars="regex",
    coords=coords,
    showdivergences=true,
)