Currently we use PyPlot. This is kind of annoying because it means that we end up depending on PyCall.jl and Conda.jl, and we need to download and install pyplot.
It would be nice to use a different Plots.jl backend. The one constraint is that the backend needs to support doing xscale = :log2. For example, the GR backend for Plots.jl does not support xscale = :log2, so we can't use it.
Currently we use PyPlot. This is kind of annoying because it means that we end up depending on PyCall.jl and Conda.jl, and we need to download and install pyplot.
It would be nice to use a different Plots.jl backend. The one constraint is that the backend needs to support doing
xscale = :log2
. For example, the GR backend for Plots.jl does not supportxscale = :log2
, so we can't use it.