NumPyCharm

Yoav Ram | Python Training

This repo contains some simple Python code to show the integation between PyCharm and Scientific Python tools:

• NumPy & matplotlib, see np_example.py:
  • set breakpoint on line 5
  • run in debug mode (green beetle)
  • step over definition of X
  • right click X in the Variables window at the bottom
  • Choose View as array to open a Data View window
  • Note how the figure window causes the run to halt; to avoid that, add the argument block=False to plt.show (but then the figure window cannot be viewed).
• Cython - support for Cython in PyCharm is limited to editing code (rather than compiling and running it) and is only available in PyCharm Professional.
  • primes.pyx defines a new Cython function that returns n first primes
  • setup.py contains neccessary definitions to build the Cython extension: use Tools->Run setup.py task..., choose build_ext and add the parameter --inplace (otherwise the Cython extension will only be available when installing the package).
  • Once the extension (pyx file) is compiled dll/so file), you can test it with test_primes.py: Right click the project name and choose Run NoseTests in ....
  • timeit_primes.py demontrates how to time both Python function (primes) and Cython functions (cprimes, imported from the primes module).
• Jupyter notebook
  • PyCharm supports viewing of Jupyter notebooks.
  • Start a notebook server by typing in a terminal (can be a PyCharm terminal): jupyter notebook
  • Note that when the notebook server starts it will display a message, look for the ...token= in the message and copy the token (the text after token).
  • Open notebook.ipynb
  • If required, paste that token you copied from the terminal.
  • You can run cells with the play button, etc.

Dependencies

• numpy
• matplotlib
• cython
• nose
• ipykernel
• jupyter notebook

License

CC-BY-SA 4.0 International