np.choose() for conditional selection in NumPy - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Predict Output

intermediate

Output of np.choose() with simple conditions

What is the output of this code snippet using np.choose() for conditional selection?

NumPy

import numpy as np
choices = [np.array([10, 20, 30]), np.array([40, 50, 60]), np.array([70, 80, 90])]
index = np.array([0, 2, 1])
result = np.choose(index, choices)
print(result)

A [10 90 50]

B [40 80 50]

C [10 80 60]

D [40 90 60]

💡 Hint

Remember that np.choose() selects elements from each array in choices based on the indices in index.

The index array has values [0, 2, 1]. For each position, np.choose() picks the element from the corresponding array in choices at that position. So, for position 0, it picks from choices[0][0] = 10; for position 1, from choices[2][1] = 80; for position 2, from choices[1][2] = 60.

❓ data_output

intermediate

Result shape and values from np.choose() with 2D arrays

Given the following code, what is the shape and content of result?

NumPy

import numpy as np
choices = [np.array([[1, 2], [3, 4]]), np.array([[5, 6], [7, 8]])]
index = np.array([[1, 0], [0, 1]])
result = np.choose(index, choices)
print(result)
print(result.shape)

A

[[1 2]
 [3 4]] with shape (2, 2)

B

[[1 6]
 [7 4]] with shape (2, 2)

C

[[5 6]
 [7 8]] with shape (2, 2)

D

[[5 2]
 [3 8]] with shape (2, 2)

💡 Hint

Each element in index selects from the corresponding position in the arrays inside choices.

For position (0,0), index=1 selects choices[1][0,0]=5; (0,1) index=0 selects choices[0][0,1]=2; (1,0) index=0 selects choices[0][1,0]=3; (1,1) index=1 selects choices[1][1,1]=8. The shape matches the input arrays (2,2).

🔧 Debug

advanced

Identify the error in np.choose() usage

What error does this code raise when run?

NumPy

import numpy as np
choices = [np.array([1, 2, 3]), np.array([4, 5])]
index = np.array([0, 1, 0])
result = np.choose(index, choices)
print(result)

A ValueError: shape mismatch: objects cannot be broadcast to a single shape

B IndexError: index 2 is out of bounds for axis 0 with size 2

C TypeError: unhashable type: 'numpy.ndarray'

D ValueError: invalid entry in choice array

💡 Hint

Check if all arrays in choices have the same shape.

The arrays in choices have different shapes: [1,2,3] has length 3, but [4,5] has length 2. np.choose() requires all arrays to have the same shape to broadcast properly, so it raises a ValueError about shape mismatch.

🚀 Application

advanced

Using np.choose() to categorize data

You have an array of scores: scores = np.array([55, 70, 85, 40, 90]). You want to assign grades: 0 for scores < 60, 1 for scores between 60 and 80, and 2 for scores above 80. Which code correctly uses np.choose() to assign grades?

A

index = np.where(scores < 60, 0, np.where(scores <= 80, 1, 2))
grades = np.choose(index, [0, 1, 2])

B

index = np.array([0 if s < 60 else 1 if s <= 80 else 2 for s in scores])
grades = np.choose(index, [np.array([0]*5), np.array([1]*5), np.array([2]*5)])

C

index = np.array([0 if s < 60 else 1 if s <= 80 else 2 for s in scores])
grades = np.choose(index, [0, 1, 2])

D

index = np.select([scores < 60, (scores >= 60) & (scores <= 80), scores > 80], [0, 1, 2])
grades = np.choose(index, [0, 1, 2])

💡 Hint

Remember that np.choose() expects arrays of the same shape in choices.

Option B creates an index array with values 0,1,2 for each score. The choices are arrays of the same shape filled with 0,1,2 respectively. np.choose() then selects the correct grade for each score. Other options misuse np.choose() by passing scalars or incorrect index arrays.

🧠 Conceptual

expert

Understanding np.choose() behavior with multidimensional index

Given choices as a list of 3 arrays each of shape (2,2), and index as a (2,2) array with values 0,1,2, what does np.choose(index, choices) return?

A An array of shape (2,2) where each element is selected from the corresponding array in choices based on index value at that position.

B A 3D array stacking all choices arrays along a new axis indexed by index array.

C A flattened 1D array combining all selected elements from choices arrays ignoring original shape.

D Raises a ValueError because index array cannot be multidimensional.

💡 Hint

Think about how np.choose() uses the index array to pick elements from each array in choices.

np.choose() uses the index array to select elements from the corresponding arrays in choices at each position. Since index and choices arrays have matching shapes, the output shape matches them, with elements chosen accordingly.