Resample with multiple columns?

[dartdog]

Using reasample and it very cool and useful.

Can I resample data with multiple columns and aggregations? for instance:

	LIST_PRICE	SOLD_PRICE
MOYRLD
1999-01-01	314900	NaN
1999-01-01	369000	369000
1999-01-01	368000	360000
1999-01-01	259900	259900
1999-01-01	239900	239900

I'd like to get both the means and counts by quarter. From the docs I see I can get the mean or count for one column eg: df3=df2.resample('Q') give the mean if I have a DF with one column, but I don't see how to do :

both(count and mean) and multiple columns?

[Jeff]

NotImplemented at this time

Its not that hard, but would need some test cases and a bit additional code

esentially you take a frame and generate a panel-like (or a multi-level indexed panel)

as a work-around as you do what you suggest: iterate over the functions and produce multiple frames

which then could be concatted together (this is pretty straightforward)

and you could actually implement it like this

If you have further interest in helping do a PR on this, pls indicate on the open issue...

and can go from there

thanks,

Jeff

https://github.com/pydata/pandas/issues/3551

[dartdog]

Thanks for the reply, was trying to figure it out, Wish I could help with implementing,

I'm still struggling to use the great features others have implemented. 

The open issue gives some good guidance.

Tom

[Jeff Reback]

np....let me know if you need more help

--
You received this message because you are subscribed to the Google Groups "PyData" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pydata+un...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.