Repository navigation
Customizable Figure layouts #315
Description
One easy thing we can start with is allowing subplots to be created with bounding boxes that described 0-100 normalized positions and widths. We can also think about more "automated" ways to define the subplots, because it can get quite tedious to write out bboxes for every subplot. The matplotlib gridspec API is kind of tedioius, we can look into other libs for inspiration as well. https://matplotlib.org/stable/gallery/lines_bars_and_markers/scatter_hist.html#defining-the-axes-positions-using-a-gridspec.
subplot1 = [bbox]
subplot2 = [bbox2]
...
gridplot = GridPlot(subplots=[subplot1, subplot2])
Activity
kushalkolar commentedon Apr 11, 2024
Some more ideas:
kushalkolar commentedon Feb 9, 2025
Some random ideas I got:
Render rectangle meshes that can be resized, use a camera to render them in screen space. Use these to set the viewport rect of subplots that are rendered on top in a second render pass. Perhaps we can render the bottom rectangle meshes only if events trigger them to resize or if the canvas resizes it has a performance impact.
Need to figure how to have an automatic mechanism so that when a rectangle is resized the adjacent rectangles resize accordingly. And move and snap behavior would be great too. See how other software that has this does it.
Axes and ticks can be within these rectangles that frame a subplot.
[-]API ideas for customizable gridplots[/-][+]Customizable Figure layouts[/+]kushalkolar commentedon Feb 16, 2025
I just realized we can't put the axes in these outer rect meshes because we have subplot docks, they will have to exist within the subplot area itself.