NumPy Sorting, Searching and Filtering
This lesson shows how to put arrays in order, find values in them and keep only the values you need.
Sorting
np.sort() returns a sorted copy. For descending order, reverse the result with [::-1].
Example
import numpy as np
prices = np.array([450, 120, 999, 75, 300])
print(np.sort(prices))
print(np.sort(prices)[::-1])
names = np.array(["Zara", "Ali", "Hina"])
print(np.sort(names))[ 75 120 300 450 999] [999 450 300 120 75] ['Ali' 'Hina' 'Zara']
np.argsort() gives the indexes that would sort the array. This is handy when you want to sort one array by another:
Example
import numpy as np
students = np.array(["Ali", "Sara", "Omar"])
marks = np.array([72, 91, 85])
order = np.argsort(marks)[::-1]
print(students[order])['Sara' 'Omar' 'Ali']
Filtering With Boolean Masks
A comparison gives an array of True / False. Put that inside the square brackets, and NumPy keeps only the True items. This is called boolean indexing or a mask.
Example
import numpy as np
marks = np.array([45, 78, 52, 30, 90, 66])
passed = marks[marks >= 50]
print(passed)
print("Number passed:", (marks >= 50).sum())[78 52 90 66] Number passed: 4
True counts as 1 when you add, so .sum() counts how many items match.
Combining Conditions
Use & for "and", the pipe symbol for "or" and ~ for "not". Put each condition in brackets.
Example
import numpy as np
ages = np.array([12, 17, 25, 34, 61, 70])
print(ages[(ages >= 18) & (ages < 60)])
print(ages[~(ages >= 18)])[25 34] [12 17]
Do not use the words and / or with arrays. They work on single True/False values only and will raise an error. Use & and the pipe symbol instead.
np.where()
np.where(condition) gives the indexes where the condition is true. With three arguments, it picks between two values item by item:
Example
import numpy as np
marks = np.array([45, 78, 52, 30])
print(np.where(marks < 50))
print(np.where(marks >= 50, "Pass", "Fail"))(array([0, 3]),) ['Fail' 'Pass' 'Pass' 'Fail']
Unique Values
Example
import numpy as np
cities = np.array(["Karachi", "Lahore", "Karachi", "Multan", "Lahore"])
values, counts = np.unique(cities, return_counts=True)
print(values)
print(counts)['Karachi' 'Lahore' 'Multan'] [2 2 1]
Joining and Splitting
| Function | What it does |
|---|---|
np.concatenate([a, b]) | Joins arrays end to end |
np.vstack([a, b]) | Stacks as rows (one on top of the other) |
np.hstack([a, b]) | Joins side by side |
np.split(a, n) | Splits into n equal parts |
Example
import numpy as np
a = np.array([1, 2, 3])
b = np.array([4, 5, 6])
print(np.concatenate([a, b]))
print(np.vstack([a, b]))
print(np.split(np.arange(6), 3))[1 2 3 4 5 6] [[1 2 3] [4 5 6]] [array([0, 1]), array([2, 3]), array([4, 5])]
Exercise
Exercise
arr = np.array([5, 12, 8, 20]). What does arr[arr > 10] return?