extract multiple ranges from a list

[pi.arctan at gmail.com]

On 8 Mar, 17:32, Peter Otten <__pete... at web.de> wrote:
> pi.arc... at gmail.com 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

Thanx guys!