This package provides a Plots.jl recipe for Sankey diagrams.
julia> ]
pkg> add SankeyPlotsusing SankeyPlots
src = [1, 1, 1, 1, 2, 2, 2, 3, 4, 5]
dst = [6, 3, 7, 4, 3, 7, 4, 7, 8, 8]
weights = [0.1, 0.3, 0.5, 0.5, 0.2, 2.8, 1, 0.45, 4.5, 3.3]
sankey(src, dst, weights)
In addition to Plots.jl attributes the following keyword arguments are supported.
| Keyword argument | Default value | Options |
|---|---|---|
node_labels |
nothing |
AbstractVector{<:String} |
node_colors |
nothing |
Vector of color specifications supported by Plots.jl or color palette |
edge_color |
:gray |
Plots.jl supported color, color selection from connected nodes with :src, :dst, :gradient, or an AbstractDict{Tuple{Int, Int}, Any} where edge_color[(src, dst)] maps to a color |
label_position |
:inside |
:legend, :node, :left, :right, :top or :bottom |
label_size |
8 |
Int |
compact |
false |
Bool |
force_layer |
Vector{Pair{Int,Int}}() |
Vectors of Int pairs specifying the layer for every node e.g. [4=>2] to force node 4 in layer 3 |
force_order |
Vector{Pair{Int,Int}}() |
Vectors of Int pairs specifying the node ordering in each layer e.g. [1=>2] to specify node 1 preceeds node 2 in the same layer |
names = [
"PV",
"Electricity Buy",
"Battery",
"Heat pump",
"Biomass",
"Electricity Sell",
"Electricity Demand",
"Heat demand",
]
energy_colors = palette(:seaborn_colorblind)[[9, 10, 3, 5, 2, 8, 1, 4]]
sankey(
src, dst, weights;
node_labels=names,
node_colors=energy_colors,
edge_color=:gradient,
label_position=:bottom,
label_size=7,
compact=true,
force_layer=[6=>2],
force_order=[5=>1]
)
@oxinabox made this possible by implementing the algorithm that solves the layout in LayeredLayouts.jl.