np.where() for conditional selection in NumPy - Cheat Sheet & Quick Revision

Recall & Review

beginner

What does np.where() do in numpy?

np.where() helps you pick values from arrays based on a condition. It returns elements chosen from two options depending on whether the condition is true or false.

beginner

How do you use np.where() to replace negative numbers in an array with zero?

You write a condition to check for negative numbers, then tell np.where() to put zero if true, else keep the original number.

np.where(arr < 0, 0, arr)

beginner

What are the three arguments of np.where(condition, x, y)?

• condition: A test that returns True or False for each element.
• x: Values to pick when condition is True.
• y: Values to pick when condition is False.

intermediate

What happens if you use np.where() with only a condition argument?

It returns the indices (positions) where the condition is True, instead of picking values.

beginner

Why is np.where() useful in data science?

It lets you quickly select or change data based on conditions without writing loops. This makes your code faster and easier to read.

What does np.where(arr > 5, 10, 0) do?

A Returns the original array unchanged

B Returns indices where elements are greater than 5

C Replaces elements less than 5 with 10, others with 0

D Replaces elements greater than 5 with 10, others with 0

np.where() picks 10 where condition is true (arr > 5), else 0.

What does np.where(condition) return if only one argument is given?

A Indices where condition is True

B Values where condition is True

C Boolean array of condition

D Error because arguments are missing

With one argument, np.where() returns indices where the condition is True.

Which of these is a correct use of np.where() to replace negative values with zero?

A np.where(arr, 0, arr)

B np.where(arr > 0, arr, 0)

C np.where(arr < 0, 0, arr)

D np.where(arr == 0, 0, arr)

Condition checks for negative values, replaces them with zero, else keeps original.

If arr = np.array([1, 2, 3]), what does np.where(arr > 2) return?

A array([False, False, True])

B (array([2]),)

C array([3])

D array([1, 2])

Returns a tuple with indices where condition is True, here index 2.

Why might np.where() be preferred over a for-loop for conditional selection?

A It is faster and uses less code

B It always uses more memory

C It only works with small arrays

D It requires writing complex loops

np.where() is vectorized, so it runs faster and is simpler than loops.

Explain how np.where() can be used to change values in an array based on a condition.

Think about how you pick between two options depending on a test.

Describe the difference between using np.where() with one argument versus three arguments.

Focus on what output you get in each case.