Not working properly with pyplot.imshow about adjusttext

[Phlya]

Hi, thanks for reporting this - interesting, I have never tried annotating an image. Can you try adding the on_basemap argument? Possibly it's the same problem there?..

…

On Fri, 15 Jun 2018, 10:33 hennumea, ***@***.***> wrote: I'm trying to annotate a picture (numpy 2D array) using pyplot.imshow, but adjust_text is not very successful. It seems to place the texts outside the axes limits nor does it handle the overlapping well I'm running Python 3.6.1, matplotlib 2.2.2, adjusttext 0.7.2. Example code: import numpy as npimport matplotlib.pyplot as pltfrom adjustText import adjust_text img = np.random.randint(low=0, high=10, size=(100, 100)) xr = np.random.randint(low=10, high=90, size=5) yr = np.random.randint(low=10, high=90, size=5) fig = plt.figure() ax = fig.add_subplot(111) ax.imshow(img, cmap='Greys', vmax='100') texts = [plt.text(x, y, s='text here', color='red', fontsize=16) for x, y in zip(xr, yr)] adjust_text(texts, arrowprops=dict(arrowstyle='-', color='red')) plt.show() — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#55>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ACwsuqW9AJpUTHDy7fIk9cvP07GVVpRaks5t839sgaJpZM4UpR2H> .

[hennumea]

I have tried switching 'on_basemap' argument, but no difference. An alternative for me is to use pyplot.pcolormesh, where adjust_text seems to work fine, but that means that I have to rewrite a good chunk of code.

pyplot.imshow places the origin of the image at the top by default, because most images (as with the ultrasound images I'm trying to annotate) are scanned from top to bottom (same with image sensors, I guess). I assume that it have something to do with how the axes are handled which confuses adjust_text.

[Phlya]

Ah, right! Inverted axes are a problem, can you just change that manually to check whether this is the problem?

…

On Fri, 15 Jun 2018, 11:46 hennumea, ***@***.***> wrote: I have tried switching 'on_basemap' argument, but no difference. An alternative for me is to use pyplot.pcolormesh, where adjust_text seems to work fine, but that means that I have to rewrite a good chunk of code. pyplot.imshow places the origin of the image at the top by default, because most images (as with the ultrasound images I'm trying to annotate) are scanned from top to bottom (same with image sensors, I guess). I assume that it have something to do with how the axes are handled which confuses adjust_text. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#55 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ACwsuh-qWgQhAlXGUgB3DbZVEkhGyk9Zks5t85B4gaJpZM4UpR2H> .

[hennumea]

When i set the origin to the bottom left corner, everything works fine. So it seems that the inverted axes are the problem, indeed.

I found a relatively quick workaround which is acceptable for me:

1. Invert the image using numpy.flip when calling imshow, and set the origin to 'lower' to avoid a reversed axis. The extent of the image alos needs to be changed (yex, 0) -> (0, yex) .
2. Invert the y-coordinate of the points to be annotated
3. Set the y-axis tick mark labels manually to display an inverted axis

More neat solutions will be appreciated!

[Phlya]

Thanks, good, then this is not a new issue, and there is a workaround for now. It should not be difficult to make it work for inverted axes, I just haven't found the time yet to do this.

…

On Fri, 15 Jun 2018, 13:26 hennumea, ***@***.***> wrote: When i set the origin to the bottom left corner, everything works fine. So it seems that the inverted axes are the problem, indeed. I found a relatively quick workaround which is acceptable for me: 1. Invert the image using numpy.flip when calling imshow, and set the origin to 'lower' to avoid a reversed axis. The extent of the image alos needs to be changed (yex, 0) -> (0, yex) . 2. Invert the y-coordinate of the points to be annotated 3. Set the y-axis tick mark labels manually to display an inverted axis More neat solutions will be appreciated! — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#55 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ACwsukmXwm0YU2MDcenofemJeiRFmfdJks5t86fkgaJpZM4UpR2H> .

[Phlya]

I'll close this since it has already been reported to multiple times - but really need to get this working!

[Phlya]

FYI the fix was even simpler than I thought, and this should now work as expected in the latest version in the repo.