[EMAIL PROTECTED] wrote:
> One of my many project involves working with YUV-files, where I need
> to reduce
> the vertical resolution with a factor of two, i.e. remove every other
> scan line.
> Today I'm using two for-loops in the fashion shown below
>
> y = []
> for i in range(0, width*height, width*2):
> for j in range(0,width):
> y.append(Y[i+j])
>
> This approach doesn't feel very pythonic but I can't come up with a
> better idea to do it.
> I've tried list comprehension and map together with lambda but I can't
> get a flattened list
> of every other scan-line...
>
> CIF = 352x288 items for luminance and the aim is to have the list
> below:
> y = [0:352 704:1056 ... ]
>>> width = 3; height = 5
>>> Y = range(width*height)
>>> y = []
>>> for i in range(0, width*height, 2*width):
... y.extend(Y[i:i+width])
...
>>> y
[0, 1, 2, 6, 7, 8, 12, 13, 14]
Probably more efficient, but needs numpy:
>>> import numpy
>>> width = 3
>>> height = 5
>>> Y = range(width*height)
>>> a = numpy.array(Y).reshape(height, width)
>>> a
array([[ 0, 1, 2],
[ 3, 4, 5],
[ 6, 7, 8],
[ 9, 10, 11],
[12, 13, 14]])
>>> b = a[::2]
>>> b
array([[ 0, 1, 2],
[ 6, 7, 8],
[12, 13, 14]])
>>> list(b.reshape(len(b)*width))
[0, 1, 2, 6, 7, 8, 12, 13, 14]
Peter
--
http://mail.python.org/mailman/listinfo/python-list
