Array Indexing
Lesson 12 of 18 in Coddy's Numpy Fundamentals course.
Another great feature in Numpy is the ability to select specific indices from an array. Place a List inside the [] brackets and specify which indices you want to select: ary[indices]
Example 1
Retrieve the 0th, 2nd, 4th elements:
x = np.arange(5, 15) # np.array([5, 6, 7, 8, 9, 10, 11, 12, 13, 14])
x[[0, 2, 4]]
>>> [5, 7, 9]Example 2
Retrieve the first and last element (but now from a 2D array)
x = np.array([[0, 1], [2, 3], [4, 5], [6, 7]])
x[[0, -1]]
>>> [[0, 1], [6, 7]]What about fetching specific elements from a 2D array?
We use ary[[Row indices], [Element indices]]
Example 3
Retrieve the first and last element from the 2nd and 3rd rows
x = np.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],
])
x[[1, 1, 2, 2], [0, -1, 0, -1]]
>>> [6, 11, 12, 17]What happened?
Numpy looks at the corresponding index of each list inside the bracket [] and returns the element:
x[[1, 1, 2, 2], [0, -1, 0, -1]]
x[1, 0] # --> 6x[1, -1] # --> 11x[2, 0] # --> 12x[2, -1] # --> 17
How to get from all rows the first and last index?
Test yourselfThis lesson includes a short quiz. Start the lesson to answer it and track your progress.
Test yourselfThis lesson includes a short quiz. Start the lesson to answer it and track your progress.
Challenge
EasyGiven a 2D array ary retrieve all numbers divided by five with no remainder (0, 5, 10, ...) with the methods we learned this lesson.
Try it yourself
import numpy as np
'''
ary looks like:
array([[ 0, 1, 2, 3, 4],
[ 5, 6, 7, 8, 9],
[10, 11, 12, 13, 14],
[15, 16, 17, 18, 19]])
'''
ary = np.arange(0, 20).reshape(4, 5) # don't change ary
res = ary[] # <-- Complete indices
print(str(res).replace("\n", "")) # don't change this lineAll lessons in Numpy Fundamentals
2N-Dimensional Array Creation
Higher DimensionUnderstanding ShapesPopulate with Fixed ValuesNumpy TypesRangeReshape3Indexing Numpy Array
Basic IndexingSlicingArray IndexingDimensional IndexingBoolean IndexingMultiple Condition Indexing