Raster scan pattern python

One possible solution, using a mix of Numpy and normal Python:

from pylab import *

# define some grids
xgrid = arange(20, 31) 
ygrid = arange(10, 16)

xscan = []
yscan = []

for i, yi in enumerate(ygrid):
    xscan.append(xgrid[::(-1)**i]) # reverse when i is odd
    yscan.append(ones_like(xgrid) * yi)   

# squeeze lists together to vectors
xscan = concatenate(xscan)
yscan = concatenate(yscan)

# quick plot
plot(xscan, yscan, '.-')
axis([19, 31, 9, 16])
show()

The way this works is to define 2 empty lists for x and y, to which you append one scan-line at a time. For the x-coordinates, you alternatively need a forward and a backward line (xgrid[::1] for forward and xgrid[::-1] for backwards, the alternating +1 and -1 are obtained by (-1)**i), while for the y-coordinate, you have to repeat a single y-value for each x-coordinate. In the end, you have to concatenate the list of vectors into a single vector.

The same can be achieved without any for-loops using a list-comprehension and repeat, like so:

xscan = concatenate([xgrid[::(-1)**i] for i in range(len(ygrid))])
yscan = repeat(ygrid, len(xgrid))

Resulting scan pattern: