[SciPy-User] embarassingly basic question

[Gustaf Nilsson]

Hi
yeah, i realise now the flaw pointed out in my question. numpy.where seems
to do exactly what i need! thanks!

Gusty

On Sun, Sep 27, 2009 at 7:37 PM, Gökhan Sever <gokhansever at gmail.com> wrote:

>
>
> On Sun, Sep 27, 2009 at 1:27 PM, Zachary Pincus <zachary.pincus at yale.edu>wrote:
>
>> Hmm, so the original question was, "if the value in the array is lower
>> than x, then make it zero, if the value is greater than x, then make
>> it 1", which requires that if the value of the array equals x, it
>> should be unchanged... Gökhan's answer does correctly and mine below
>> does not.
>>
>> However, this is  a bit of an unusual request, more often one would
>> want to make everything zero if it's < x and 1 if >= x (or <= and >,
>> respectively). In this case, the answers below work, but there's an
>> even simpler special-case answer for zeros and ones:
>>
>> a >= x
>>
>> gives a boolean array with zeros where a < x and ones where a >= x.
>> (All basic logic operations work here.)
>>
>> Zach
>>
>>
>>
> I saw that in my posting, too; leaving the comparison point intact.
>
> np.where seems a more elegant approach to me if greater or less than
> comparison what Gustaf was asking for. It is a one-liner in the end.
> Element-wise functionality of numpy is really so powerful and practical. The
> same approach doesn't work on regular Python sequences, or could they be
> forced to work similar to numpy's arrays?
>
>
>
>>
>>
>> On Sep 27, 2009, at 2:18 PM, Zachary Pincus wrote:
>>
>> > numpy.where is also useful for this case:
>> >
>> > In : numpy.where(numpy.arange(10) < 5, 0, 1)
>> > Out: array([0, 0, 0, 0, 0, 1, 1, 1, 1, 1])
>> >
>> > Note that there's a lot going on in this example, including
>> > broadcasting the scalar 0 and 1 values to 1D arrays... where can take
>> > any array with a "compatible" shape as the second or third argument:
>> > In : numpy.where(numpy.arange(10) < 5, numpy.arange(20, 30),
>> > numpy.arange(60, 70))
>> > Out: array([20, 21, 22, 23, 24, 65, 66, 67, 68, 69])
>> >
>> >
>> > On Sep 27, 2009, at 1:53 PM, Gökhan Sever wrote:
>> >
>> >> On Sun, Sep 27, 2009 at 12:44 PM, Gustaf Nilsson <
>> gustaf at laserpanda.com
>> >>> wrote:
>> >> Hi
>> >> (this might even be a numpy question, not scipy)
>> >> how do i do conditionals numpy arrays?
>> >> What i want to do is: if any value in the array is lower than x,
>> >> then make it zero, if the value is greater than x, then make it 1
>> >>
>> >> i tried googling the answer, but dont think i used the right keywords
>> >>
>> >> cheers
>> >> Gusty
>> >>
>> >>
>> >>
>> >> Robert Kern must be sleeping :)
>> >>
>> >> I[1]: a = arange(10)
>> >>
>> >> I[2]: a[a<5] = 0
>> >>
>> >> I[3]: a
>> >> O[3]: array([0, 0, 0, 0, 0, 5, 6, 7, 8, 9])
>> >>
>> >> I[4]: a[a>5] = 1
>> >>
>> >> I[5]: a
>> >> O[5]: array([0, 0, 0, 0, 0, 5, 1, 1, 1, 1])
>> >>
>> >>
>> >> --
>> >> ■ ■ ■ ■ ■ ■ ■ ■ ■ ■
>> >>
>> >> _______________________________________________
>> >> SciPy-User mailing list
>> >> SciPy-User at scipy.org
>> >> http://mail.scipy.org/mailman/listinfo/scipy-user
>> >>
>> >>
>> >>
>> >>
>> >> --
>> >> Gökhan
>> >> _______________________________________________
>> >> SciPy-User mailing list
>> >> SciPy-User at scipy.org
>> >> http://mail.scipy.org/mailman/listinfo/scipy-user
>> >
>> > _______________________________________________
>> > SciPy-User mailing list
>> > SciPy-User at scipy.org
>> > http://mail.scipy.org/mailman/listinfo/scipy-user
>>
>> _______________________________________________
>> SciPy-User mailing list
>> SciPy-User at scipy.org
>> http://mail.scipy.org/mailman/listinfo/scipy-user
>>
>
>
>
> --
> Gökhan
>
> _______________________________________________
> SciPy-User mailing list
> SciPy-User at scipy.org
> http://mail.scipy.org/mailman/listinfo/scipy-user
>
>


-- 
■ ■ ■ ■ ■ ■ ■ ■ ■ ■
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.scipy.org/pipermail/scipy-user/attachments/20090927/34c447ac/attachment.html>