qqplot and qqnorm

qqplot(x, y; kwargs...)

Draw a Q-Q plot, comparing quantiles of two distributions. y must be a list of samples, i.e., AbstractVector{<:Real}, whereas x can be

  • a list of samples,

  • an abstract distribution, e.g. Normal(0, 1),

  • a distribution type, e.g. Normal.

In the last case, the distribution type is fitted to the data y.

The attribute qqline (defaults to :none) determines how to compute a fit line for the Q-Q plot. Possible values are the following.

  • :identity draws the identity line.

  • :fit computes a least squares line fit of the quantile pairs.

  • :fitrobust computes the line that passes through the first and third quartiles of the distributions.

  • :none omits drawing the line.

Broadly speaking, qqline = :identity is useful to see if x and y follow the same distribution, whereas qqline = :fit and qqline = :fitrobust are useful to see if the distribution of y can be obtained from the distribution of x via an affine transformation.

Graphical attributes are

  • color to control color of both line and markers (if markercolor is not specified)

  • linestyle

  • linewidth

  • markercolor

  • strokecolor

  • strokewidth

  • marker

  • markersize

qqnorm(y; kwargs...)

Shorthand for qqplot(Normal(0,1), y), i.e., draw a Q-Q plot of y against the standard normal distribution. See qqplot for more details.

Examples

Test if xs and ys follow the same distribution.

using CairoMakie

xs = randn(100)
ys = randn(100)

qqplot(xs, ys, qqline = :identity)

Test if ys is normally distributed.

using CairoMakie

ys = 2 .* randn(100) .+ 3

qqnorm(ys, qqline = :fitrobust)