imcube_t = tensor(imcube_bw)
df = pd.DataFrame(imcube_t[1:32,1:32])
df.style.set_properties(**{'font-size':'6pt'}).background_gradient('Greys')It seems the PILImageBW.create, will also revert the colors. Is that correct?
Placing 255-tensor(imcube_bw) is a nice workaround, but still not sure, what I was doing wrong.
imcube_t = 255-tensor(imcube_bw)
df = pd.DataFrame(imcube_t[1:32,1:32])
df.style.set_properties(**{'font-size':'6pt'}).background_gradient('Greys')
Yep, that’s the default behaviour, you can do:
PILImageBW._show_args['cmap'] = 'gray'
to change that
Or you can create your custom type like described here
1 Like
Thank you so much Lucas, is there a specific computational reason why Jeremy used ImageBW and not the cmap to gray?
I have also noticed we have IntToFloatTensor in Chapt 11. Great to use with DataLoaders and pipelines