np.full() for custom-filled arrays in NumPy - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ Predict Output

intermediate

Output of np.full() with integer fill value

What is the output of this code?

import numpy as np
arr = np.full((2, 3), 7)
print(arr)

NumPy

import numpy as np
arr = np.full((2, 3), 7)
print(arr)

A

[[7 7 7]
 [7 7 7]]

B

[[7 7 7 7]
 [7 7 7 7]]

C

[[7 7]
 [7 7]
 [7 7]]

D

[[0 0 0]
 [0 0 0]]

💡 Hint

Look at the shape argument (2, 3) and the fill value 7.

np.full creates an array of the given shape filled with the fill value. Here, shape (2,3) means 2 rows and 3 columns filled with 7.

❓ data_output

intermediate

Shape and dtype of np.full() array

What is the shape and data type of the array created by this code?

import numpy as np
arr = np.full((4,), 3.14, dtype=np.float32)
print(arr.shape, arr.dtype)

NumPy

import numpy as np
arr = np.full((4,), 3.14, dtype=np.float32)
print(arr.shape, arr.dtype)

A (4,) int64

B (4, 1) float64

C (1, 4) float32

D (4,) float32

💡 Hint

Check the shape tuple and the dtype argument.

The shape is (4,) meaning a 1D array with 4 elements. The dtype is float32 as specified.

🔧 Debug

advanced

Identify the error in np.full() usage

What error does this code raise?

import numpy as np
arr = np.full(3, [1, 2])
print(arr)

NumPy

import numpy as np
arr = np.full(3, [1, 2])
print(arr)

A ValueError: could not broadcast input array from shape (2,) into shape ()

B No error, prints array with shape (3, 2)

C SyntaxError: invalid syntax

D TypeError: fill_value must be a scalar

💡 Hint

np.full expects a scalar fill value, but here a list is given.

np.full expects a scalar fill value. Passing a list causes a broadcasting error because it cannot fit the list into each scalar element.

🚀 Application

advanced

Creating a masked array with np.full()

You want to create a 3x3 numpy array filled with -1 to mark missing data. Which code correctly creates this array with integer type?

A np.full((3, 3), '-1')

B np.full((3, 3), -1, dtype=int)

C np.full([3, 3], -1.0, dtype=int)

D np.full((3, 3), -1, dtype='float')

💡 Hint

Check the shape format and dtype for integer type.

Option B uses a tuple for shape, integer fill value, and dtype=int which is correct. Option B uses string fill value, C uses list for shape which is allowed but less common, D uses float dtype which is not integer.

🧠 Conceptual

expert

Memory behavior of np.full() with mutable fill value

What happens if you run this code?

import numpy as np
arr = np.full((2,), [0])
arr[0].append(1)
print(arr)

NumPy

import numpy as np
arr = np.full((2,), [0])
arr[0].append(1)
print(arr)

A Raises TypeError because fill_value is a list

B First element is [0, 1], second is [0] because lists are copied

C Both elements show [0, 1] because the same list object is used

D Prints [[0] [0]] unchanged

💡 Hint

Think about how np.full handles mutable objects as fill_value.

np.full uses the same object reference for each element if fill_value is mutable. So modifying one element modifies all references.