TensorFlow Tensors

Hands-on practice for tensor creation, shapes, math operations, and manipulation.

Creating Tensors

import tensorflow as tf

scalar = tf.constant(7)
vector = tf.constant([10, 10])
matrix = tf.constant([[1,2],[3,4]])

tensor = tf.constant([
 [[1,2],[3,4]],
 [[5,6],[7,8]]
])

Key idea:
Rank = number of dimensions
Shape = size of each dimension

tf.Variable vs tf.constant

changeable = tf.Variable([10,7])
unchangeable = tf.constant([10,7])

changeable[0].assign(7)   # Works
# unchangeable[0].assign(7)  Error

Random Tensors

random_1 = tf.random.Generator.from_seed(42).normal((3,2))
random_2 = tf.random.Generator.from_seed(42).normal((3,2))

Tensor Shapes & Indexing

rank_4 = tf.zeros([2,3,4,5])

rank_4.shape
rank_4.ndim
tf.size(rank_4)

rank_4[:1,:1,:1,:]

Math Operations

tensor = tf.constant([[10,7],[3,4]])

tensor + 10
tensor * 10
tf.matmul(tensor, tensor)

Aggregation

E = tf.constant(np.random.randint(0,100,50))

tf.reduce_min(E)
tf.reduce_max(E)
tf.reduce_mean(E)
tf.reduce_sum(E)

One-Hot Encoding

indices = [0,1,2,3]
tf.one_hot(indices, depth=4)

Exercises

1. Create scalar, vector, matrix, tensor
2. Find shape, rank, size
3. Create random tensors [5,300]
4. Matrix multiply them
5. One-hot encode a tensor