Question

I have a set of matrices as numpy 2d arrays in a list. They are of different sizes (19×19 to 63×63).. Now, I have to arrange them starting from the one with the smallest size to the largest. I am inputting the matrices as below:

pics=[]
templates=[]

c1=0
for c1 in range(1,31):
    i1=Image.open("data/"+str(c1)+".gif")
    pmat = numpy.asarray(i1)
    pics.append(i1)
    templates.append(pmat)

Thus, templates is the list that has all the image matrices. pmat of course, contains the images themselves. Is there any easy way to sort this?….

Answer 1

You can sort a list of images by length, see below:

>>> import numpy 
>>> a = numpy.array([1,2,3])
>>> b = numpy.array([1,3])
>>> c = [a,b]
[array([1, 2, 3]), array([1, 3])]
>>> sorted(c,key=len)
[array([1, 3]), array([1, 2, 3])]

If your image arrays are all square (ie. x-len = y-len) then this should work no problem for your purposes.

EDIT: If you need more flexibility (for non-square images) you can use key=numpy.size.