DOC: comparison between Datafeames.jl, pandas, and r data.fame objects

[sglyon]

It seems like a lot of the functionality of this package is inspired by things in R data.frames or pandas. I think more users would be willing to give this package a try if in the documentation there was a section on the differences between dataframes and the others. Maybe even a table serving as a sort of cross reference between the three and how to perform common routines using each language.

This is somewhat reminiscent of the section on notable differences between julia and matlab, R, and python from the julia documentation and the old blog post about R and julia syntax from @johnmyleswhite .

[johnmyleswhite]

If you could help contribute such a table, it would be very much appreciated.

[milktrader]

I think this would be cool to have @spencerlyon2

[sglyon]

I'm happy to work on this for pandas + julia, but I am somewhat inexperienced when it comes to R (at least with respect to the other people who have worked on this project), so I'd ask for some help in filling in the R "column"

[milktrader]

Yeah, I could help there. What's the plan, start a branch on your fork maybe?

[sglyon]

Sure, that sounds good. I'll start it up and post here once I have it read to go.

Thanks

[ViralBShah]

This would certainly be very useful to have.

[sglyon]

Before I get too far on this, how are the docs built right now? I see that they are markdown, but am not sure which variant?

[sglyon]

Also, I haven't worked a ton with DataFrames in Julia so if anyone has suggestions for examples of non trivial things I should include in the table just let me know. Volunteering to do this was as much a way of forcing myself to get more comfortable with this package, as it was anything altruistic.
:)

I'll probably be pinging this issue occasionally to get input.

[milktrader]

You can start with non-trivial things in Pandas and then fill in the blanks.

[sglyon]

Good idea, that will definitely be my approach, but there are bound to be some Julia specific things (like using quote to specify the DataFrame one column at a time) that I probably haven't come across yet.

[sglyon]

Very very rough draft here. Have't done much besides create the file

[sglyon]

Sorry guys but it turns out I will have less time in the near future than I thought. I will probably not get this done any time soon.

If someone else wants to take the lead, feel free. If you do, let me know and I'll try to contribute.

[StefanKarpinski]

No worries. Thanks for getting it started.

[milktrader]

You can leave the branch open on your fork and I'll drop some more additions into it.

[sglyon]

Sounds good, I really haven't done much of anything, but I did give you push access to my fork.

[milktrader]

I already added some R syntax and I'll fill some more stuff in when I get a chance.

[quinnj]

I'm not sure there's enough interest in this; anyone feel free to re-open if still desired.

[nalimilan]

I think that would be useful, and the number of comments by different people seems to confirm this.

EDIT: #793 contains a file presenting a comparison with R and Pandas. http://pandas.pydata.org/pandas-docs/stable/comparison_with_r.html can also give some inspiration.

[bkamins]

Closing this as we will soon have this (we already have it for dplyr and Stata)

[nalimilan]

Almost 7 years later at last! :-)

[bkamins]

Well - I would say that we are now finally ready for this 😄.

Just need to finish #2340 and I am ready for PyData Global 2020 tutorial how to port from pandas to DataFrames.jl that is upcoming in November this year.