Plot the confidence interval for a model fit

Specific Domains Statistics
, ,
4

EDIT: I accidentally led this thread down a rabbit hole by plotting a prediction interval rather than a confidence interval below, and also bungling the use of ribbon. Long story short the ggplot2 plot shown below can also be obtained in Plots.jl/GLM.jl if one calls predict(model, pred, interval = :confidence, level = 0.95).

@BLI has the right plot command, let me add the GLM command you are probably after:


using DataFrames, GLM, Plots
data = DataFrame(x = rand(100));
data.y = 1 .+ 2*data.x .+ 0.1*rand(100);

model = lm(@formula(y ~ x), data)
pred = DataFrame(x = 0:0.01:1);
pr = predict(model, pred, interval = :prediction, level = 0.95)

plot(xlabel="x", ylabel="y", legend=:bottomright)
plot!(data.x, data.y, label="data", seriestype=:scatter)
plot!(pred.x, pr.prediction, label="model", linewidth=3, 
        ribbon = (pr.prediction .- pr.lower, pr.upper .- pr.prediction))

image
image812×549 35.3 KB

The relevant docstring is:

?help> predict


  predict(mm::LinearModel, newx::AbstractMatrix;
          interval::Union{Symbol,Nothing} = nothing, level::Real = 0.95)

  If interval is nothing (the default), return a vector with the predicted values for model mm and new data newx. Otherwise, return a 3-column matrix with the prediction and the lower and upper confidence bounds for a given level (0.95 equates alpha = 0.05).
  Valid values of interval are :confidence delimiting the uncertainty of the predicted relationship, and :prediction delimiting estimated bounds for new data points.
5 Likes