CODEWITHSUNDEEP
pythonperformanceoptimizationlist-comprehension
Gianni Spear
Gianni Spear

Reputation: 7980

Python: use of List Comprehensions in order to fill an empty list single object-by-single object

i have a 'dictionary-valueiterator' Groups. For each list of Groups i random sample two (or more values) and i store in a new list

GroupsList = list([])
  for m in Groups:
     GroupsList.append(point_random_selection(m,n))

where point_random_selection

def point_random_selection(list,n):
    if n == 0:
            raise TypeError("The pulse density 0 is not iterable")
    try:
        sample_point = random.sample(list,n)
    except ValueError:
        sample_point = list
    return sample_point

GroupsList = [[<liblas.point.Point object at 0x00000001363A0E10>],
 [<liblas.point.Point object at 0x00000002CE52CB38>],
 [<liblas.point.Point object at 0x00000000CF516908>],
 [<liblas.point.Point object at 0x00000004293F9400>],
 [<liblas.point.Point object at 0x0000000249F67C50>],
 [<liblas.point.Point object at 0x0000000312D85A90>],
 [<liblas.point.Point object at 0x000000047DF396A0>,
  <liblas.point.Point object at 0x000000047DF14F98>]]

where in some case i have two or more list (ex:)

GroupsList[6]
[<liblas.point.Point object at 0x000000047DF396A0>,
 <liblas.point.Point object at 0x000000047DF14F98>]

I wish to store one list-by-one list in the new list in order to append one-by-one also when i have two or more liblas.point.Point object as:

GroupsList = [[<liblas.point.Point object at 0x00000001363A0E10>],
     [<liblas.point.Point object at 0x00000002CE52CB38>],
     [<liblas.point.Point object at 0x00000000CF516908>],
     [<liblas.point.Point object at 0x00000004293F9400>],
     [<liblas.point.Point object at 0x0000000249F67C50>],
     [<liblas.point.Point object at 0x0000000312D85A90>],
     [<liblas.point.Point object at 0x000000047DF396A0>],
     [<liblas.point.Point object at 0x000000047DF14F98>]]

Is it a List Comprehensions the best approach? ex to insert in 

GroupsList.append(point_random_selection(m,n))

Upvotes: 0

Views: 167

Answers (2)

Andrew Clark
Andrew Clark

Reputation: 208625

If I understand this correctly, you want to change a nested list like [[a, b], [c, d]] into [a, b, c, d] using a list comprehension. If that is the correct interpretation for what you are trying to do, the following will work:

GroupsList = [point for lst in GroupsList for point in lst]

However it would be better to just create the flat structure from the beginning by changing your loop to the following (replacing append() with extend()):

for m in Groups:
    GroupsList.extend(point_random_selection(m,n))

a.extend(b) (where a and b are lists) results in the same behavior as the following loop (but does it more efficiently):

for x in b:
    a.append(x)

Upvotes: 2

Danica
Danica

Reputation: 28856

I'm not entirely sure I'm reading your question correctly, but are you looking for list.extend? For example:

>>> a = [1, 2]
>>> a.extend([5, 1, 4])
>>> a
[1, 2, 5, 1, 4]

Also, some general style feedback:

  • Standard Python style strongly discourages naming variables things like Group or GroupsList; name them groups or groups_list.
  • list([]) is more easily written as just []
  • Naming the argument to point_random_selection list should probably be avoided, since it shadows the global list() function
  • It seems like you could replace most of point_random_selection by
return random.sample(the_list, min(n, len(the_list))

Upvotes: 1

Related Questions