[R] Running sum

Douglas Bates bates at stat.wisc.edu
Sat Nov 20 01:00:26 CET 2004


Marc Schwartz wrote:
> On Fri, 2004-11-19 at 16:44 -0600, hadley wickham wrote:
> 
>>Perhaps I'm missing something, but isn't the maximum of the cumulative
>>sum simply the last value, ie. sum(x)?
>>
>>Hadley
> 
> 
> 
> Indeed!
> 
> And...going back to Sean's original post he did write:
> 
> "I just need max(Y)."
> 
> Thus, a whole group of us (independently) managed to over-engineer the
> solution. 
> 
> Presuming of course, that Sean does not require the entire cumulative
> sum vector for other purposes.
> 
> I wonder what that says about human cognition....
> 
> Thanks,
> 
> Marc
> <Cleaning his bi-focals...>

I have forgotten the details of the original question but it seems to me 
that if the elements of x could be both positive and negative then the 
maximum of the cumulative sum doesn't have to be the last sum.




More information about the R-help mailing list