Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.lang.python > #35145

Re: calculation on lists

Date 2012-12-19 12:24 -0600
From Tim Chase <python.list@tim.thechases.com>
Subject Re: calculation on lists
References <CAHzaPEPXMLSqhqyzSj96xFcedjr=HeoPv+d_xGrOwXAqLe-PFQ@mail.gmail.com> <BLU168-W6060C53B0CD9109AF1D0068F300@phx.gbl>
Newsgroups comp.lang.python
Message-ID <mailman.1066.1355941422.29569.python-list@python.org> (permalink)

Show all headers | View raw


On 12/19/12 09:24, loïc Lauréote wrote:

>> is there a tool to calculate on list ?
>>
>> something like :
>>
>>> a= [1,1,1,1]
>>> b = [5,9,8,4]
>>> c = a+b*a
>>> print c
>>> [6,10,9,5]
>>
>> Thx
>>
>> ======
>>
>> Hi,
>> for such simpler cases, you may try list comprehensions and probably
>> the zip(...) function
>>
>>>>> [a+b*a for a,b in zip([1,1,1,1], [5,9,8,4])]
>> [6, 10, 9, 5]
> Thank for your answer,
> 
> I found something allowing to avoid loops. I use operator
> overloading.

This is a pretty obvious case of "don't duplicate others' effort".
With stock python, Vlastimil's list-comprehension using zip() is the
canonical answer.  I'm pretty sure it's far faster than your class,
in addition to feeling far more pythonic.

If you happen to have numpy installed, Vlastimil's answer there
already does what you want, but has the added benefits of

1) being heavily tested

2) having core parts written in C for efficiency

3) having a complete intuitive interface, rather than yours which
may have issues (such as not having left-vs-right issues:

  a * 5
  5 * a

should produce the same results)

4) both the list-comprehension version and the numpy version aren't
limited to 2-element vectors like your code, but can handle
N-element vectors.

-tkc


Back to comp.lang.python | Previous | Next | Find similar | Unroll thread


Thread

Re: calculation on lists Tim Chase <python.list@tim.thechases.com> - 2012-12-19 12:24 -0600

csiph-web