Closed
Description
Bug summary
The set_val() method of a range slider doesn't set the value correctly with values close to the minimal and maximal values of the range slider. With values in the middle, everything works fine.
Code for reproduction
import matplotlib.pyplot as pyplot
import matplotlib.widgets as widgets
ax = pyplot.axes([0.15, 0.01, 0.7, 0.1])
ax2 = pyplot.axes([0.15, 0.21, 0.7, 0.1])
ax3 = pyplot.axes([0.15, 0.41, 0.7, 0.1])
ax4 = pyplot.axes([0.15, 0.61, 0.7, 0.1])
# correct behaviour with set_val using values in the middle between val min and val max
correct2 = widgets.RangeSlider(ax, "correct2", valmin=1.0, valmax=10.0)
correct2.set_val((4, 6))
print(correct2.val)
# correct with val init
correct = widgets.RangeSlider(ax2, "correct", valinit=(1.0, 2.0), valmin=1.0, valmax=10.0)
print(correct.val)
# wrong with set_val having values close to the value max
wrong2 = widgets.RangeSlider(ax3, "wrong2", valmin=1.0, valmax=10.0)
wrong2.set_val((9, 10))
print(wrong2.val)
# wrong with set_val having values close to the value min
wrong = widgets.RangeSlider(ax4, "wrong", valmin=1.0, valmax=10.0)
wrong.set_val((1, 2))
print(wrong.val)
pyplot.show()Actual outcome
The values for the rangesliders "wrong" and "wrong2" are not set correctly
Expected values are: (1, 2) and (9, 10)
Actual values are (1, 3.25) and (7.75, 10)
Expected outcome
The result of using set_val should be the same, as when using the valinit parameter of the constructor.
Additional information
The problem also occurred on Manjaro with:
- Python version: 3.10.9
- Matplotlib version: 3.6.2
- Matplotlib backend: QtAgg
- Installation of matplotlib via Linux package manager
Operating system
Windows 10
Matplotlib Version
3.6.2
Matplotlib Backend
TkAgg
Python version
3.11.0
Jupyter version
No response
Installation
pip