Multiple Category Axis

[alecloudenback]

Feature Request: Plotting with different groups/layers on the axis.

https://peltiertech.com/chart-with-a-dual-category-axis/

[Mattriks]

For the plot linked above, you can do:

using DataFrames, Gadfly
groups =  vcat(fill.(["Mechanical", "Electrical", "Hydraulic"], [3,5,3])...)
components = ["Gear", "Bearing", "Motor", "Electrical\nSwitch", "Plug", "Cord", "Fuse", 
    "Bulb", "Hydraulic\nPump", "Leak", "Seals"]
rates = [11, 8, 3, 19, 12, 11, 3, 2, 4, 3, 1]
df = DataFrame(group = groups, component=components, rate=rates)

p1 = plot(df, x=:component, y=:rate, color=:group, Geom.bar, Theme(bar_spacing=1mm),
    Guide.xlabel("Mechanical\t\t\t\tElectrical\t\t\tHydraulic  "))

If the "components" were the same for the different groups, try Geom.subplot_grid. Also example here.

[bjarthur]

@alecloudenback if the example above solves your problem can you please close this issue? thanks.

[alecloudenback]

Thanks! I appreciate getting it to work on the one example linked, but I thought that since the workaround was very manual that it would still be nice to have better support for this type of plotting built in.

[Mattriks]

Here's what currently happens with Geom.subplot_grid(free_x_axis=true):

# df same as above
plot(df, x=:component, y=:rate, xgroup=:group, color=:group, 
    Geom.subplot_grid(Geom.bar, free_x_axis=true),
    Theme(bar_spacing=1mm), Guide.xlabel(nothing)
    )

The issue is that the subplot x scales are not proportional (so the bars are of different width). Thus this is the same issue as #1093.