-
Notifications
You must be signed in to change notification settings - Fork 13
Open
Description
The following script should result in a plot where the text labels, as well as the line segments, are cropped to be within the viewport.
import vcs
x = vcs.init(geometry=(600, 600))
vp = [0.25, 0.75, 0.1, 0.9]
xs = [0.0, 1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0]
ys = [0.2, 0.4, 0.6, 0.5, 0.7, 0.3, 0.2, 0.1, 0.9, 0.4]
txt = [str(idx) for idx in range(len(xs))]
xmin = min(xs)
xmax = max(xs)
ymin = min(xs)
ymax = max(ys)
dx = 2
l = x.createline()
l.color = ['Green',]
l.viewport = vp
l.worldcoordinate = [xmin + dx, xmax - dx, ymin, ymax]
l.x = xs
l.y = ys
t = x.createtext()
t.viewport = vp
t.worldcoordinate = [xmin + dx, xmax - dx, ymin, ymax]
t.color = 'Red'
t.x = xs
t.y = ys
t.string = txt
t.height = 10
x.plot(l)
x.plot(t)
x.png('texts-viewport-issue.png')However, the lines are cropped within the viewport but texts are not, as can be seen in this image:
I do not think this issue was introduced by the update to Context2D rendering in the VTK backend, because the thermo repo has an old baseline which shows that the problem has existed all along, but `vcs didn't have a test that caught it.
I have a fix in a branch on my fork, here. However, that branch illustrates that there is a problem with the templated viewport coordinates, as in projected plots, the viewport isn't big enough to allow room for the tick labels.
Metadata
Metadata
Assignees
Labels
No labels
