Pass arguments to Secondary Axis scaling functions?

Development
1

Axes.secondary_xaxis takes as an argument a tuple of functions that convert between the two axis systems. Is it possible to pass arguments along to these functions when I call secondary_xaxis?

Here is my use case:
I have a function setup_plot that sets the axes, legend, etc of a plot. The plot will be given a primary linear x axis and a secondary x axis that scales by (unique_param / x). That unique_param is different for each plot, so I would like to be able to call setup_plot(axis, unique_param), and pass that param on to the scaling function when it calls axis.secondary_xaxis().

2

I suggest using https://docs.python.org/3.8/library/functools.html#functools.partial to close over your parameter, so something like


def setup_plot(ax, unique_param):
    def forward(T, param):
       return T * param

    def inv(d, param):
        return d / param

    ax.secondary_xaxis(
        -0.2, 
        functions=(partial(forward, param=unique_param),
                   partial(inv, param=unique_param))
     )
3

Thank you! This takes care of getting the unique parameter into the scaling function’s namespace. I didn’t know that one could define a function within a function in python. I tried this while omitting the partial and it seemed to work fine; what is the reason for adding the partial()?

Something like this would be a nice example for the Secondary Axis figure gallery. I think my situation might be a common one: adding an axis scale that depends on a tunable parameter of the experiment.

4

I was sleepy and did not think too hard about where I defined the conversion functions.

partial is more-or-less doing the same thing as defining a function in the function (and creating a closure), but in a slightly terser way and doing a bit more book keeping (if you look at the repr for example the partial one will be a bit nicer).