Documentation: Vaex / Numpy equivalent commands

[DougRzz]

In the documentation, it would be v useful to have a list / table of Numpy /Pandas / Dask commands alongside the equivalent Vaex command.

Suggested commands could include:

• numpy.histogram(a, array) | vx.count(....., binby(a,) )
• numpy.2dhistogram | vx.count(....., binby(a,b) )
• numpy.histogram(.... , weights = weightsArray )
• numpy.diff
• pandas.rolling_mean
• numpy.cumsum
• numpy.fft

(N.B, all the commands above are supported by Dask)

There could be many more commands listed. There may be no equivalent Vaex command but it would be useful if the documentation could state either:

• Planned for release #.# (see github issue ###)
• help wanted
• no plans to support
• no direct equivalent command but here is a hack.......

[maartenbreddels]

I like the idea, since for some parts vaex works slightly different. I'm currently working on a refactoring of the project, code and docs, and I'm gonna keep this in mind.
I am not sure all functions make sense, for instance, to do an fft you should just use numpy I think. Would a howto section in the docs help, to answer these questions?

[DougRzz]

Hi Maarten,

Sorry for the late reply. I had a think about a 'howto' section in the docs. Then I thought perhaps a Jupyter notebook would work well if the tutorial of Vaex/Numpy equivalent commands was somehow interactive. I have finally got round to creating a Notebook which includes a few Vaex commands but no Numpy equivalents yet. The Python code for each action is printed below the widgets, so hopefully, this would enable the user to more easily connect between the GUI and the Vaex Python commands. Hopefully it might be useful to some Vaex beginners. If you get the time to review it, please let me know what you think.

https://github.com/DougRzz/Vaex-demo-notebook/blob/master/VAEX_interactive.ipynb