Python | Tensorflow log1p() method

Last Updated : 18 Jan, 2022

Tensorflow is an open-source machine learning library developed by Google. One of its applications is to develop deep neural networks.
The module tensorflow.math provides support for many basic mathematical operations. Function tf.log1p() [alias tf.math.log1p] provides support for the natural logarithmic function in Tensorflow. It expects the input in form of complex numbers as $a+bi$ or floating point numbers. The input type is tensor and if the input contains more than one element, an element-wise logarithm of $1+x$ is computed, $y=\log_e (1+x)}$$ .

Syntax: tf.log1p(x, name=None) or tf.math.log1p(x, name=None)
Parameters:
x: A Tensor of type bfloat16, half, float32, float64, complex64 or complex128.
name (optional): The name for the operation.
Return type: A Tensor with the same size and type as that of x.

Code #1:

Python3

# Importing the Tensorflow library
import tensorflow as tf
 
# A constant vector of size 5
a = tf.constant([-1.5, -1, -0.5, 0, 0.5, 1, 1.5], dtype = tf.float32)
 
# Applying the log1p function and
# storing the result in 'b'
b = tf.log1p(a, name ='log1p')
 
# Initiating a Tensorflow session
with tf.Session() as sess:
    print('Input type:', a)
    print('Input:', sess.run(a))
    print('Return type:', b)
    print('Output:', sess.run(b))

Output:

Input type: Tensor("Const:0", shape=(7, ), dtype=float32)
Input: [-1.5 -1.  -0.5  0.   0.5  1.   1.5]
Return type: Tensor("log1p:0", shape=(7, ), dtype=float32)
Output: [        nan        -inf -0.6931472   0.          0.4054651   0.6931472
  0.91629076]

$nan$ denotes that natural logarithm of 1+x doesn’t exist for negative values and $-inf$ denotes that it approaches to negative infinity as the input approaches to -1.
Code #2: Visualization

Python3

# Importing the Tensorflow library
import tensorflow as tf
 
# Importing the NumPy library
import numpy as np
 
# Importing the matplotlib.pyplot function
import matplotlib.pyplot as plt
 
# A vector of size 20 with values from -1 to 0 and 0 to 10
a = np.append(np.linspace(-1, 0, 10), np.linspace(0, 10, 10))
 
# Applying the logarithmic function and
# storing the result in 'b'
b = tf.log1p(a, name ='log1p')
 
# Initiating a Tensorflow session
with tf.Session() as sess:
    print('Input:', a)
    print('Output:', sess.run(b))
    plt.plot(a, sess.run(b), color = 'red', marker = "o") 
    plt.title("tensorflow.abs") 
    plt.xlabel("X") 
    plt.ylabel("Y") 
    plt.grid()
 
    plt.show()

Output:

Input: [-1.         -0.88888889 -0.77777778 -0.66666667 -0.55555556 -0.44444444
 -0.33333333 -0.22222222 -0.11111111  0.          0.          1.11111111
  2.22222222  3.33333333  4.44444444  5.55555556  6.66666667  7.77777778
  8.88888889 10.        ]
Output: [       -inf -2.19722458 -1.5040774  -1.09861229 -0.81093022 -0.58778666
 -0.40546511 -0.25131443 -0.11778304  0.          0.          0.7472144
  1.17007125  1.46633707  1.69459572  1.88031287  2.03688193  2.17222328
  2.29141179  2.39789527]