Fix: adapted grid to work with Number

[briochemc]

This PR simply extends the type of the (min, max)-tuple argument for the adaptive_grid to accept Numbers instead of just Reals. The goal is to fix jw3126/UnitfulRecipes.jl#36 🙂.

For example, with this PR, this MWE

using Unitful: m
using Plots, UnitfulRecipes
f(x) = sin(x / m)
plot(f, -5m, 5m)

would give

which otherwise currently throws a

MethodError: no method matching adapted_grid(..., ::Tuple{Quantity, Quantity})

on master.

Hopefully, this PR does not break anything since it's a tiny edit! (I have not run tests locally so waiting to see the CI runs... 😁)

[briochemc]

Well... Every CI workflow failed, so I guess this is not as simple a fix I thought! 😅

[briochemc]

I did not realize that the types of min and max could be different, which is why the CI failed before, so I just fixed that in the latest commit. Hopefully this works! 🤞

[briochemc]

The CI fails seem on these zscale tests:

PlotUtils.jl/test/runtests.jl

Lines 183 to 189 in dd8042c

	if VERSION ≥ v"1.5"
	@test cmin ≈ 2827.440 atol = 1e-3
	@test cmax ≈ 4172.194 atol = 1e-3
	else
	@test cmin ≈ 2823.715 atol = 1e-3
	@test cmax ≈ 4185.414 atol = 1e-3
	end

which seems unrelated to this PR. BTW, since these seem to be Julia-version dependent (no idea why), they might need to be updated for Julia v1.6, v1.7, and so on?

[briochemc]

Just in case, cc @mileslucas for the zscale CI failures? 🤷

[mileslucas]

test failures for zscale are due to ever-changing random number generation 😞. I'll open a separate issue to triage this- I don't think it should block this PR.

[briochemc]

@daschw maybe this can be reviewed for merging before the next release? (Saying this because I saw there was PR #109.)

[daschw]

Nice! Thanks a lot!