Cannot retrieve the version information of OpenCV

[vmirly]

I can import cv2 package in my jupyter notebook, but watermark cannot find the proper information for this package:

DistributionNotFound: The 'cv2' distribution was not found and is required by the application

[rasbt]

Hey, Vahid,

hm, that's weird. Can you import the 'cv2' package in your current jupyter session via
__import__('cv2') ?

[vmirly]

Sebastian,

Yes, importing via this command works in jupyter notebook:

__import__('cv2')
<module 'cv2' from '/home/vahid/.virtualenvs/cv/lib/python3.4/site-packages/cv2.so'>

Also, the version can be seen via

cv2.__version__
3.1.0'

[rasbt]

Hm, what watermark version are you using then? If you are using a version < v. 1.3.2 you may have to use the install name ("opencv") instead of the import package name.

PS: you can get the watermark version via %watermark -w

[vmirly]

Something is not right. I successfully installed the latest version (1.3.4)

  Found existing installation: watermark 1.3.4
    Uninstalling watermark-1.3.4:
      Successfully uninstalled watermark-1.3.4
  Running setup.py develop for watermark
Successfully installed watermark

but in jupyter I still get the old version:

import watermark
%watermark -w
watermark v. 1.2.3

[rasbt]

Hm, this sounds more like a general issue related to where/how you installed packages -- maybe sth related to your virtual environments? Maybe you installed a second version of the package in an env, and your jupyter comes from another environment? Maybe just upgrade the package just from Jupyter notebook directly like

%%sh pip install watermark --upgrade

[rasbt]

PS: it looks like you ran python setup.py develop? This does not install the package but it will create an .egg-link in the deployment directory back to the project source code directory.

[vmirly]

Now it seems that the problem is resolved, and %watermark -v -p cv2 gives the OpenCV version.

I had tried pip install watermark --upgrade as well, but for some reasons, it was not working.

[rasbt]

Hm, I suspect there was something weird about the virtual env that you upgraded and the virtual env that loaded the notebook. However, I am glad that it's sorted out now :)