[Numpy-discussion] calculating the difference of an array

[Manuel Wittchen]

Hi,

Thanks for your help.
I tried np.diff() before, but the result looked like this:

RESULT = [1, 1, 1, 1]

So I was thinking that np.diff() doesn't iterate over the values of
the array. So I gave the for-loop a try.
Now, seeing your code below, I realized that my mistake was that I
used ARRAY = [0, 1, 2, 3, 4, 5] for the calculations...

Stupid me.


2010/1/2 Emmanuelle Gouillart <emmanuelle.gouillart@normalesup.org>:
> Hello Manuel,
>
> the discrete difference of a numpy array can be written in a very
> natural way, without loops. Below are two possible ways to do it:
>>>> a = np.arange(10)**2
>>>> a
> array([ 0,  1,  4,  9, 16, 25, 36, 49, 64, 81])
>>>> a[1:] - a[:-1]
> array([ 1,  3,  5,  7,  9, 11, 13, 15, 17])
>>>> np.diff(a) # another way to calculate the difference
> array([ 1,  3,  5,  7,  9, 11, 13, 15, 17])
>
> The error in the example you give is due to the fact that you iterate
> over len(ARRAY), which is an integer, hence not an iterable object. You
> should write ``for i in range(len(ARRAY))`` instead.
>
> Cheers,
>
> Emmanuelle
>
> On Sat, Jan 02, 2010 at 11:23:13AM +0100, Manuel Wittchen wrote:
>> Hi,
>
>> I want to calculate the difference between the values of a
>> numpy-array. The formula is:
>
>> deltaT = t_(n+1) - t_(n)
>
>> My approach to calculate the difference looks like this:
>
>> for i in len(ARRAY):
>>       delta_t[i] = ARRAY[(i+1):] - ARRAY[:(len(ARRAY)-1)]
>
>> print "result:", delta_t
>
>> But I get a TypeError:
>> File "./test.py", line 19, in <module>
>>     for i in len(ARRAY):
>> TypeError: 'int' object is not iterable
>
>> Where is the mistake in the code?
>
>> Regards and a happy new year,
>> Manuel Wittchen
>> _______________________________________________
>> NumPy-Discussion mailing list
>> NumPy-Discussion@scipy.org
>> http://mail.scipy.org/mailman/listinfo/numpy-discussion
> _______________________________________________
> NumPy-Discussion mailing list
> NumPy-Discussion@scipy.org
> http://mail.scipy.org/mailman/listinfo/numpy-discussion
>