Uh oh!
There was an error while loading.Please reload this page.
- Notifications
You must be signed in to change notification settings - Fork7.9k
Closed
Description
I found some rather weird behavior (maybe a bug) when trying to useaxes.transData.transform((1,1)) with a log axis. The basic code to reproduce the error is below
#!/usr/bin/env python3from matplotlib.backends.backend_agg import FigureCanvasAgg as FigureCanvasfrom matplotlib.figure import Figureimport numpy as npfig = Figure(figsize=(8,6))canvas = FigureCanvas(fig)ax = fig.add_subplot(1,1,1)ax.set_yscale('log') # <--- works fine without this lineax.transData.transform((1,1)) # <--- exception thrown hereThe odd thing is that it seems to work fine if I givetransform a 1x2 shaped array, i.e.
ax.transData.transform(np.array((1,1)).reshape(1,-1))Unfortunately, this means the examples in thetransformations tutorial don't work with a log axis.