best cumulative sum

[Bengt Richter]

On Mon, 21 Nov 2005 15:23:20 +1100, Steven D'Aprano <steve at REMOVEMEcyber.com.au> wrote:

>David (Alan) Isaac wrote:
>
>> What's the good way to produce a cumulative sum?
>> E.g., given the list x,
>> cumx = x[:]
>> for i in range(1,len(x)):
>>  cumx[i] = cumx[i]+cumx[i-1]
>> 
>> What's the better way?
>
>Is there something that this doesn't do, or something 
>it does do that it shouldn't?
>
>You could do it this way:
>
># untested
>def cumulative_sum(L):
>     CL = []
>     csum = 0
>     for x in L:
>         csum += x
>         CL.append(csum)
>     return CL
>
>Whether it is better or worse depends on what you 
>consider better or worse.

I think x[:] ought to be a faster space allocation than building by appending,
but I think there is too much indexing in the OP's version,
so I think I would combine ideas, e.g.,

 # (still ;-) untested
 def cumulative_sum(L):
      CL = L[:]
      csum = 0
      for i, x in enumerate(L):
          csum += x
          CL[i] = csum
      return CL

I minor nit might be in how/when promotions to long or float happen if L has
mixed types. The OP didn't mention any requirements, but e.g. his version won't ever
promote the first output element, since he walks replacements from L[1] on.

( posting delayed >12 hrs due to news server prob ;-/ )

Regards,
Bengt Richter