Type casting in TensorFlow - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to cast a tensor to float32.

TensorFlow

import tensorflow as tf
x = tf.constant([1, 2, 3])
y = tf.cast(x, [1])

Drag options to blanks, or click blank then click option'

Atf.float32

Btf.int64

Ctf.bool

Dtf.string

Attempts:

3 left

💡 Hint

Common Mistakes

Using tf.int64 will keep the tensor as integers.

Using tf.bool or tf.string will cause errors or unexpected results.

tf.cast converts the tensor to the specified data type. Here, tf.float32 is the correct type for floating point numbers.

2fill in blank

medium

Complete the code to cast a float tensor to int32.

TensorFlow

import tensorflow as tf
x = tf.constant([1.5, 2.7, 3.9])
y = tf.cast(x, [1])

Drag options to blanks, or click blank then click option'

Atf.float64

Btf.int32

Ctf.bool

Dtf.string

Attempts:

3 left

💡 Hint

Common Mistakes

Using tf.float64 keeps the tensor as floats.

Using tf.bool or tf.string causes errors.

tf.int32 converts the float tensor to integers by truncating the decimal part.

3fill in blank

hard

Fix the error in casting a boolean tensor to float32.

TensorFlow

import tensorflow as tf
x = tf.constant([True, False, True])
y = tf.cast(x, [1])

Drag options to blanks, or click blank then click option'

Atf.int32

Btf.string

Ctf.float32

Dtf.bool

Attempts:

3 left

💡 Hint

Common Mistakes

Using tf.string causes runtime errors.

Using tf.bool does not change the type.

Boolean tensors can be cast to float32 to get 1.0 for True and 0.0 for False.

4fill in blank

hard

Fill both blanks to cast a tensor of integers to boolean and then back to int64.

TensorFlow

import tensorflow as tf
x = tf.constant([0, 1, 2, 0])
y = tf.cast(x, [1])
z = tf.cast(y, [2])

Drag options to blanks, or click blank then click option'

Atf.bool

Btf.float32

Ctf.int64

Dtf.string

Attempts:

3 left

💡 Hint

Common Mistakes

Casting directly to tf.string causes errors.

Casting to tf.float32 is not needed here.

First cast integers to boolean (0 becomes False, others True), then cast back to int64 (False=0, True=1).

5fill in blank

hard

Fill all three blanks to create a tensor, cast it to float64, then to int32, and finally to boolean.

TensorFlow

import tensorflow as tf
x = tf.constant([10, 0, 5])
y = tf.cast(x, [1])
z = tf.cast(y, [2])
w = tf.cast(z, [3])

Drag options to blanks, or click blank then click option'

Atf.float64

Btf.int32

Ctf.bool

Dtf.string

Attempts:

3 left

💡 Hint

Common Mistakes

Casting to tf.string causes errors.

Changing the order of casts may cause unexpected results.

The tensor is first cast to float64, then to int32 (truncating decimals if any), and finally to boolean (0 becomes False, others True).