NumPy Indexing and Slicing
Indexing picks single values from an array, and slicing picks a range of values.
Indexing a 1-D Array
Indexes start at 0, just like Python lists. Negative indexes count from the end.
Example
import numpy as np
temps = np.array([31, 33, 35, 34, 30, 29, 32])
print(temps[0])
print(temps[3])
print(temps[-1])
print(temps[1] + temps[2])31 34 32 68
Indexing a 2-D Array
Give two numbers separated by a comma: row first, then column.
Example
import numpy as np
# rows = students, columns = Maths, English, Science
marks = np.array([
[85, 72, 90],
[60, 88, 75],
[92, 95, 89]
])
print(marks[0, 2]) # first student, Science
print(marks[2, 0]) # third student, Maths
print(marks[1]) # the whole second row
print(marks[-1, -1]) # last row, last column90 92 [60 88 75] 89
Slicing a 1-D Array
The pattern is [start:stop:step]. The stop index is not included.
Example
import numpy as np
arr = np.array([10, 20, 30, 40, 50, 60, 70])
print(arr[1:4])
print(arr[:3])
print(arr[4:])
print(arr[::2])
print(arr[::-1])[20 30 40] [10 20 30] [50 60 70] [10 30 50 70] [70 60 50 40 30 20 10]
| Slice | Meaning |
|---|---|
arr[1:4] | Index 1, 2, 3 |
arr[:3] | From the start up to index 2 |
arr[4:] | From index 4 to the end |
arr[::2] | Every second item |
arr[::-1] | All items in reverse |
Slicing a 2-D Array
Slice rows and columns separately, with a comma between them. A single : means "all".
Example
import numpy as np
marks = np.array([
[85, 72, 90],
[60, 88, 75],
[92, 95, 89]
])
print(marks[:, 0]) # Maths column for all students
print(marks[0:2, 1:]) # first two students, English and Science[85 60 92] [[72 90] [88 75]]
Changing Values
You can change one value or a whole slice at once:
Example
import numpy as np
arr = np.array([1, 2, 3, 4, 5])
arr[0] = 100
arr[2:4] = 0
print(arr)[100 2 0 0 5]
Slices Are Views
A slice of a NumPy array is a view: it shares data with the original. If you change the slice, the original changes too. Use .copy() when you want a separate array.
Example
import numpy as np
arr = np.array([1, 2, 3, 4])
view = arr[0:2]
copy = arr[0:2].copy()
view[0] = 99
copy[1] = 55
print(arr)[99 2 3 4]
Changing view changed arr. Changing copy did not.
Fancy Indexing
You can pass a list of indexes to pick several items in any order:
Example
import numpy as np
arr = np.array([5, 10, 15, 20, 25])
print(arr[[0, 2, 4]])[ 5 15 25]
Exercise
Exercise
For a 2-D array m, what does m[:, 1] return?