search

JuliaML / MLPlots.jl

Plotting recipes for statistics and machine learning using Plots.jl
Other
24 stars 5 forks source link

readme

MLPlots (deprecated)

Note: this package is out of date and not usable with newer versions of Julia.

Build Status

Common plotting recipes for statistics and machine learning.

This package uses RecipesBase to provide mappings from types defined in statistics and machine learning packages to generic descriptions of visualization attributes and data. These recipes can then be used in conjunction with Plots.jl to provide flexible statistical and machine learning visualizations which are independent of both the platform and graphical library.

Many recipes are conditionally included and are loaded on the relevant using call of the library. For example using ROCAnalysis, MLPlots will load plotting recipes for roc curves defined in ROCAnalysis. Recipes include:

Status: This package is currently on hiatus until further notice. Collaboration is welcomed and encouraged!

OnlineAI

Neural nets with OnlineAI.jl. Show the current state of a neural net:

using OnlineAI, MLPlots
net = buildClassificationNet(3, 1, [15,10,5])
plot(net)

onlineai1

For spiking neuron models, a spike (or raster) plot is useful to see firing times among neurons:

n = 20
spikes = SpikeTrains(n, title = "Spike Trains", color = :darkblue)
for t=1:100, i=1:n
    if rand() < 0.1
        push!(spikes, i, t)
    end
end
spikes.plt

onlineai

ROCAnalysis

ROCAnalysis.jl

using ROCAnalysis, MLPlots
curve = ROCAnalysis.roc(2+2randn(1000), -2+2randn(100000))
plot(curve)

rocanalysis