Extension of PythonPlot that allows for easy plotting with common Julia packages (Unitful.jl, Measurements.jl)
This package extends common PythonPlot.jl functions for common Julia packages with specialized data types. For instance, this allows for simple shortcuts to plot DimArrays (from DimensionalData.jl). It also allows for plotting of Measurements (data types with values and uncertainties from Measurements.jl) and Quantities (values with units from Unitful.jl). I attempt to follow typical matplotlib formatting conventions, but for some edge cases (e.g., a ribbon plot with a central estimate and uncertainty), new syntax must be used to distinguish between the different matplotlib elements being used (e.g, a plot and a fill_between).
Here, we show an example of three different extensions of PythonPlot that allow for easy plotting of DimArrays, as plain values, with units, and with uncertainties.
data = randn(3, 10)K
dataunc = abs.(randn(3,10))K
ax1 = Symbol.(["A", "B", "C"])
ax2 = 6yr:1yr:15yr
da = DimArray(data .± dataunc, (X(ax1), Ti(ax2)))
figure()
N = 3
subplot(N, 1, 1)
plot(ustrip.(value.(da[At(:A), :])))
title("DimArray, with no units, no uncertainty")
subplot(N,1,2)
plot(value.(da[At(:A), :]), lwcentral = 5, alpha = 1, color = "purple")
title("DimArray, with units, no uncertainty")
subplot(N, 1, 3)
plot(da[At(:A), :])
title("DimArray, with units, with uncertainty")
xlabel("Time [years CE]", fontsize = 15)
tight_layout()
savefig("../plots/dimarray.png")
Other functionality includes ignoring Missing values while plotting. This package is not exhaustive for every possible combination of parameters. Future functionality could include extending other common plotting recipes, like contourf, contour, etc...