From: Ryan on 2 Nov 2009 11:42
I am a long time user of "R" for statistical analysis and find it an
extremely useful environment for data exploration. That said I may be
adding and/or moving to Python for more of my work and wanted to know
if people had any recommendations.
My work is primarily financial time series analysis. In R this means
extensive use of "zoo" and "xts". I see from searching that there is a
time-series package for Python called pytseries (http://
pytseries.sourceforge.net/) that looks closest to adding the time
series capabilities to Python.
My question is to people who have used zoo and xts in R, what has
their experience been with Python? Would you recommend using native
Python to do time series analysis, or rPy to link back to R for
From: Ishwor Gurung on 2 Nov 2009 12:02
2009/11/3 Ryan <rsheftel(a)gmail.com>:
> I am a long time user of "R" for statistical analysis and find it an
> My question is to people who have used zoo and xts in R, what has
> their experience been with Python? Would you recommend using native
> Python to do time series analysis, or rPy to link back to R for
Although I haven't used 'zoo' nor 'xts' in R I use Rpy/Rpy2 because it
seems to expose almost all(iirc) the necessary R space within Python
space. So, if I need features in R - for e.g., importing modules -
library('foo'), I can still have it in Python via the Rpy/Rpy2 bridge.
On the other hand with pytseries Python library, I'd be limited to
only run time-series analysis wouldn't I? In the end though, it all
depends on your project requirements, resources and so forth..