When slicing in Python the third parameter is the step. As others mentioned, see Extended Slices for a nice overview.
With this knowledge, [::3]
just means that you have not specified any start or end indices for your slice. Since you have specified a step, 3
, this will take every third entry of something
starting at the first index. For example:
>>> '123123123'[::3]
'111'
Did I miss or nobody mentioned reversing with [::-1]
here?
# Operating System List
systems = ['Windows', 'macOS', 'Linux']
print('Original List:', systems)
# Reversing a list
#Syntax: reversed_list = systems[start:stop:step]
reversed_list = systems[::-1]
# updated list
print('Updated List:', reversed_list)
source: https://www.programiz.com/python-programming/methods/list/reverse
Python sequence slice addresses can be written as a[start:end:step] and any of start, stop or end can be dropped. a[::3]
is every third element of the sequence.
You can also use this notation in your own custom classes to make it do whatever you want
class C(object):
def __getitem__(self, k):
return k
# Single argument is passed directly.
assert C()[0] == 0
# Multiple indices generate a tuple.
assert C()[0, 1] == (0, 1)
# Slice notation generates a slice object.
assert C()[1:2:3] == slice(1, 2, 3)
# If you omit any part of the slice notation, it becomes None.
assert C()[:] == slice(None, None, None)
assert C()[::] == slice(None, None, None)
assert C()[1::] == slice(1, None, None)
assert C()[:2:] == slice(None, 2, None)
assert C()[::3] == slice(None, None, 3)
# Tuple with a slice object:
assert C()[:, 1] == (slice(None, None, None), 1)
# Ellipsis class object.
assert C()[...] == Ellipsis
We can then open up slice objects as:
s = slice(1, 2, 3)
assert s.start == 1
assert s.stop == 2
assert s.step == 3
This is notably used in Numpy to slice multi-dimensional arrays in any direction.
Of course, any sane API should use ::3
with the usual "every 3" semantic.
The related Ellipsis
is covered further at: What does the Ellipsis object do?
s[i:j:k]
is, according to the documentation, "slice of s from i to j with step k". When i
and j
are absent, the whole sequence is assumed and thus s[::k]
means "every k-th item".
First, let's initialize a list:
>>> s = range(20)
>>> s
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]
Let's take every 3rd item from s
:
>>> s[::3]
[0, 3, 6, 9, 12, 15, 18]
Let's take every 3rd item from s[2:]
:
>>> s[2:]
[2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]
>>> s[2::3]
[2, 5, 8, 11, 14, 17]
Let's take every 3rd item from s[5:12]
:
>>> s[5:12]
[5, 6, 7, 8, 9, 10, 11]
>>> s[5:12:3]
[5, 8, 11]
Let's take every 3rd item from s[:10]
:
>>> s[:10]
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
>>> s[:10:3]
[0, 3, 6, 9]
seq[::n]
is a sequence of each n
-th item in the entire sequence.
Example:
>>> range(10)[::2]
[0, 2, 4, 6, 8]
The syntax is:
seq[start:end:step]
So you can do (in Python 2):
>>> range(100)[5:18:2]
[5, 7, 9, 11, 13, 15, 17]
The third parameter is the step. So [::3] would return every 3rd element of the list/string.
This visual example will show you how to a neatly select elements in a NumPy Matrix (2 dimensional array) in a pretty entertaining way (I promise). Step 2 below illustrate the usage of that "double colons" ::
in question.
(Caution: this is a NumPy array specific example with the aim of illustrating the a use case of "double colons" ::
for jumping of elements in multiple axes. This example does not cover native Python data structures like List
).
Say we have a NumPy matrix that looks like this:
In [1]: import numpy as np
In [2]: X = np.arange(100).reshape(10,10)
In [3]: X
Out[3]:
array([[ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9],
[10, 11, 12, 13, 14, 15, 16, 17, 18, 19],
[20, 21, 22, 23, 24, 25, 26, 27, 28, 29],
[30, 31, 32, 33, 34, 35, 36, 37, 38, 39],
[40, 41, 42, 43, 44, 45, 46, 47, 48, 49],
[50, 51, 52, 53, 54, 55, 56, 57, 58, 59],
[60, 61, 62, 63, 64, 65, 66, 67, 68, 69],
[70, 71, 72, 73, 74, 75, 76, 77, 78, 79],
[80, 81, 82, 83, 84, 85, 86, 87, 88, 89],
[90, 91, 92, 93, 94, 95, 96, 97, 98, 99]])
Say for some reason, your boss wants you to select the following elements:
"But How???"... Read on! (We can do this in a 2-step approach)
Specify the "start index" and "end index" in both row-wise and column-wise directions.
In code:
In [5]: X2 = X[2:9,3:8]
In [6]: X2
Out[6]:
array([[23, 24, 25, 26, 27],
[33, 34, 35, 36, 37],
[43, 44, 45, 46, 47],
[53, 54, 55, 56, 57],
[63, 64, 65, 66, 67],
[73, 74, 75, 76, 77],
[83, 84, 85, 86, 87]])
Notice now we've just obtained our subset, with the use of simple start and end indexing technique. Next up, how to do that "jumping"... (read on!)
We can now specify the "jump steps" in both row-wise and column-wise directions (to select elements in a "jumping" way) like this:
In code (note the double colons):
In [7]: X3 = X2[::3, ::2]
In [8]: X3
Out[8]:
array([[23, 25, 27],
[53, 55, 57],
[83, 85, 87]])
We have just selected all the elements as required! :)
Now we know the concept, we can easily combine step 1 and step 2 into one consolidated step - for compactness:
In [9]: X4 = X[2:9,3:8][::3,::2]
In [10]: X4
Out[10]:
array([[23, 25, 27],
[53, 55, 57],
[83, 85, 87]])
Done!
Python uses the :: to separate the End, the Start, and the Step value.
Source: Stackoverflow.com