Optimized fast image for linspace coords #393
Conversation
… ratio which happens with imshow
| @@ -0,0 +1,249 @@ | |||
| # SPDX-License-Identifier: BSD-3-Clause | |||
There was a problem hiding this comment.
This file was moved basically without changes
| **kwargs, | ||
| if all( | ||
| (data.coords[dim].ndim < 2) | ||
| and ((data.coords[dim].dtype == str) or (sc.islinspace(data.coords[dim]))) |
There was a problem hiding this comment.
This requires that there are dim-coords for each dim. (Because it uses data.coords[dim].) Was this a requirement already before?
There was a problem hiding this comment.
Yes. If the input is missing them, some dummy (arange) coords are added before you get to here.
There are tests for this. See e.g.
plopp/tests/plotting/plot_1d_test.py
Line 51 in 69a60a3
| rasterized: bool = True, | ||
| **kwargs, | ||
| if all( | ||
| (data.coords[dim].ndim < 2) |
There was a problem hiding this comment.
Can FastImage handle scalar coords?
There was a problem hiding this comment.
If there is a scalar coord that does not have the same name as a dimension
da = pp.data.data2d()
da.coords['scalar'] = sc.scalar(333., unit='K')
da.plot()is fine (the 'scalar' coord is just ignored.
Interestingly, if the scalar coord is named 'x'
da = pp.data.data2d()
da.coords['x'] = sc.scalar(333., unit='K')
dait fails, but also fails on main.
I can fix in another PR...
This adds two branches for plotting images: the old pcolormesh that can handle any coords, and a fast optimized branch that uses imshow to show images with linspace coords.
Fixes #392