[Python Edinburgh] Python for Data Analysis?

Mike Mineter m.mineter at ed.ac.uk
Mon Dec 22 12:58:56 CET 2014


To comment on the R or Python question....

As in Rob's case, the choice can be less to do with fundamental  
preference for R or Python and more to do with what libraries are  
already available for the kind of work to be done; and with what other  
libraries  I need to work.

A bridge does exist with RPy2 - I've only used that with simple R  
calculations being called from Python but success depended on a)  
having R installed with the necessary shared libraries being built, b)  
the correct version of readline. On one machine it worked easily; on  
another I decided just to translate small R functions to python  
instead of fighting the installations to get RPy2 to go.

Mike



Quoting Douglas Houston <douglasrhouston at googlemail.com> on Mon, 22  
Dec 2014 10:16:35 +0000:

> Oh no, not come across Pandas before, I'll add it to the list (sigh)...
>
> You're right about learning R being a struggle by the way.
>
> On 22 December 2014 at 08:07, Rob Schneider <rmschne at rmschneider.com> wrote:
>
>> Have this book (electronic version) and used it a lot. It's really about
>> Pandas (http://pandas.pydata.org) which is to me very interesting.
>>
>> There are lots of web articles you can find of R vs. Pandas vs. Python and
>> all that.
>>
>> For me, bottom line, is that I can plug in Pandas to the large Python app
>> I work with and use Pandas/Matplotlib to do a small bit of number crunching
>> with it ... rather than struggle to learn and integrate R.  As I see what
>> Pandas can do I'll only use more of it as time moves on.
>>
>> --rms
>>
>>
>> On 22 Dec 2014, at 08:01, Douglas Houston <douglasrhouston at googlemail.com>
>> wrote:
>>
>> > Sorry, forgot the Coursera details:
>> >
>> > http://i.imgur.com/joL2LdF.jpg
>> >
>> > On 22 December 2014 at 08:00, Douglas Houston <
>> douglasrhouston at googlemail.com> wrote:
>> > Does anyone have this book?
>> >
>> >
>> http://www.amazon.co.uk/Python-Data-Analysis-Wrangling-IPython/dp/1449319793
>> >
>> > I'm thinking about my planned future use of Python and what I want to do
>> with it.  I'm looking to get into data science and using it to train
>> machine learning algorithms. The following shows the Coursera courses I've
>> signed up to (10 in total) to learn how to do this.
>> >
>> >
>> > Should I use Python (in which case a book like this looks useful), or
>> should I use R?
>> >
>> > My R and Python skills are probably on a par right now, but I'm unlikely
>> to have the time to get properly fluent in both at the same time any time
>> soon, so I have to make a decision.
>> >
>> > The Data Science Coursera course is focussed solely on R, so perhaps I
>> should just roll with that?
>> >
>> > Any one have any advice?
>> >
>> > _______________________________________________
>> > Edinburgh mailing list
>> > Edinburgh at python.org
>> > https://mail.python.org/mailman/listinfo/edinburgh
>>
>> _______________________________________________
>> Edinburgh mailing list
>> Edinburgh at python.org
>> https://mail.python.org/mailman/listinfo/edinburgh
>>
>



-- 
________________________________________

Dr Mike Mineter
Room 352, Grant Institute
0131 651 9094


-- 
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.




More information about the Edinburgh mailing list