[Numpy-discussion] transform an array of points efficiently?

[Chris Colbert]

no, because dot(x,y) != dot(y,x)

>>> x = np.random.rand(3,4)
>>> y = np.random.rand(4,4)
>>> np.dot(x, y)
array([[ 1.67624043,  1.66719374,  1.72465017,  1.20372021],
       [ 0.70046162,  0.60187869,  0.73094349,  0.4604766 ],
       [ 0.78707401,  1.01959666,  0.61617829,  0.43147398]])
>>> np.dot(y, x[0,:
... ])
array([ 1.44627767,  1.09332339,  2.66001046,  1.13972652])
>>> np.dot(y, x[1,:])
array([ 0.27854715,  0.56261048,  0.7793413 ,  0.44260709])
>>> np.dot(y, x[2,:])
array([ 0.70468211,  0.42843143,  1.34022702,  0.53021987])
>>>


hence I need xnew = [Transform]*[xold]

and not [xold]*[Transform]




On Thu, Jul 9, 2009 at 10:22 PM, Keith Goodman<kwgoodman@gmail.com> wrote:
> On Thu, Jul 9, 2009 at 7:08 PM, Chris Colbert<sccolbert@gmail.com> wrote:
>> say i have an Nx4 array of points and I want to dot every [n, :] 1x4
>> slice with a 4x4 matrix.
>>
>> Currently I am using apply_along_axis in the following manner:
>>
>> def func(slice, mat):
>>     return np.dot(mat, slice)
>>
>> np.apply_along_axis(func, arr, 1, mat)
>>
>> Is there a more efficient way of doing this that doesn't require a
>> python function for each slice?
>
> I'm sure I'm missing an important point, but can't you solve the whole
> problem with one dot:
>
>>> x = np.random.rand(3,4)
>>> y = np.random.rand(4,4)
>
>>> np.dot(x, y)
>
> array([[ 0.86488057,  0.23456114,  0.91592677,  0.89798689],
>       [ 1.24197754,  0.39907686,  1.45453141,  1.13645076],
>       [ 1.41419289,  0.81818818,  1.09768428,  1.32719635]])
>
>>> np.dot(x[0,:], y)
>   array([ 0.86488057,  0.23456114,  0.91592677,  0.89798689])
>>> np.dot(x[1,:], y)
>   array([ 1.24197754,  0.39907686,  1.45453141,  1.13645076])
>>> np.dot(x[2,:], y)
>   array([ 1.41419289,  0.81818818,  1.09768428,  1.32719635])
> _______________________________________________
> NumPy-Discussion mailing list
> NumPy-Discussion@scipy.org
> http://mail.scipy.org/mailman/listinfo/numpy-discussion
>