docker-compose.override.yml does not allow matplotlib visualizations from Python Console

Hi

I have the following issue:

When I use docker-compose on my own ( without pycharm), both matplotlib's plt.show() and cv2.show() work fine. 

When I use  docker-compose as remote interpreter from console with Pycharm19 profeccional, matplotlib's  plt.show() stops  ploting the result without any error/warning poping up and cv2.imshow() works fine as before. 

I suppose the problem is  in .PyCharm2019.1/system/tmp/docker-compose.override.<some_number>.yml file, which overrides/adds additional Pycharm configs in docker-compose that breaks down matplotlib module. I have several questions:

1. How I can overcome this issue if I want to use Pycharm from it's console

2. Can I configure pycharm's docker-compose.override file ? 

3. If my assumption about docker-compose.override.yml  is wrong, then what can be the problem of such strange behaviour ?

Additionally, if I Python Console  from Pycharm with docker but without docker-compose , can I set running configurations ( i.e. docker run ...custom running configurations...)? 

Thanks in advance

1 comment

Do you have anything specific in mind that breaks matplotlib in docker-compose.override? 

It shouldn't be modified under normal circumstances, so I would rather reproduce the issue and report to developers. Do you have a simplified example you can show?

0

Please sign in to leave a comment.