0

I get this error when trying to import gensim on Jupyter notebook launched via Ananconda. error.

I've checked and gensim is definitely installed. Also made sure that numpy is installed. Not sure why the error is showing up. Can someone please help me identify what the error is and suggest how to fix it.

  • 2
    write your error not put the picture – Ilya Aug 15 '23 at 06:45
  • Please provide enough code so others can better understand or reproduce the problem. – Community Aug 15 '23 at 12:57
  • On the one hand, like other SO contributors, I'm frustrated when askers supply a screenshot-of-text, instead of copying & pasting text itself – which should be just as easy, & speeds all sorts of later reading/analysis, by people & systems. But on the other hand, OCR is now trivial – especially for monospace fonts dominant in asker screenshots. Here in 2023, shouldn't SO's submission interface detect screenshots-of-text & offer to replace or decorate such images with full `PRE` text? Or even do so automatically? If SO staff devs don't have time for this, can they ask ChatGPT to implement? – gojomo Aug 15 '23 at 16:17

1 Answers1

0

You're getting the same error message as is described here: ValueError: numpy.ndarray size changed, may indicate binary incompatibility. Expected 88 from C header, got 80 from PyObject

So, the approaches suggested there may help: upgrading or uninstalling-then-reinstalling numpy, or any other packages installed (such as gensim here).

(If you'd pasted in your error test, rather than the screenshot, StackOverflow's own "related questions" suggestions may have automatically pointed you at those answers, or other similar ones.)

If all else fails, you might want to try starting again from a fresh minimal conda environment, installing just what you need step-by-step, to ensure all the packages are recent & mutually-compatible versions.

gojomo
  • 52,260
  • 14
  • 86
  • 115