Newbie : innerproduct function from Numarray

[Diez B. Roggisch]

> 1)
> from Numarray import innerproduct
> output = []
> temp = vector1 + vector1    # temp is twice the length of vector1
> for i in range(2000):
>      output.append(innerproduct(temp[i:(i+2000)],vector2)
> 2)
> output = []
> temp = vector1 + vector1
> for i in range(2000):
>     sum = 0
>     for j in range(2000):
>         sum += temp[i+j] * vector2[j]
>     output.append(sum)
> 
> I thought the first method using Numarray should be faster.
> But it looks like the second method is faster.
> Am I doing anything wrong?
> Do you guys know any faster way to do this?

First of all, I assume that both results are equal :)

>From the numarray-docs:

----
innerproduct(a, b) 
innerproduct produces the inner product of arrays a and b. It is equivalent
to matrixmultiply(a, transpose(b)). 
----

I'm not sure what this means mathematically (I understand the operation
made, but I'm not sure what an inner product _means_).

However, what you do with your hand-written code doesn't look like what I
would write if I would come up with my own implementation of the
aforementioned definition of innerproduct. Matrix-multiplication is
O(n**3), while your code is in O(n**2). So it seems that your special-case
with vectors, not arrays, produces an easier to compute variant of
innerproduct - and the differenes of n-times is of course important.

BTW: Its better to use xrange instead of range, it won't create the actual
list of numbers, but an iterable object instead - saves memory and time :)

Diez