Python TensorFlow truncated normal

In this Python tutorial, we will learn how to use the truncated normal function in TensorFlow. Additionally, we will cover the following topics.

• TensorFlow distribution truncated_normal
• TensorFlow truncated_normal_initailizer
• TensorFlow truncated normal example
• TensorFlow has no attribute ‘truncated_normal’
• TensorFlow has no attribute ‘truncated_normal_initializer’

Python TensorFlow Truncated normal

• In this section, we will discuss how to use the truncated normal function in Python TensorFlow.
• To perform this particular task, we are going to use the tf.random.truncated_normal() function and this method is used to generate random values from a normal distribution and normal distribution means probability distribution that occurs in many events.
• In this example, we are going to use a normal constant variable that will consider as the shape of the tensor and apply in tf.random.truncated_normal() function.

Syntax:

Let’s have a look at the Syntax and understand the working of tf.random.truncated_normal() function.

tf.random.truncated_normal
                          (
                           shape,
                           mean=0.0,
                           stddev=1.0,
                           dtype=tf.dtypes.float32,
                           seed=None,
                           name=None
                          )

• It consists of a few parameters
  • shape: This parameter indicates the output of the given tensor and it must be a one-dimensional Tensor that contains an only integer value.
  • mean: By default it takes 0 value and it specifies that we have to find the mean value of the given normal distribution.
  • stddev: This parameter indicates that we have to find the standard deviation of the given normal distribution and by default, it takes a 1.0 value.
  • dtype: By default it takes tf.dtypes.float32() value and it indicates the datatype of output.
  • name: This is an optional parameter and defines the name of the operation(truncate name).

Example:

Let’s take an example and check how to use the tf.random.truncated_normal() function.

Source Code:

import tensorflow as tf

tf.compat.v1.disable_eager_execution()
tens = tf.constant(10,dtype="int32",name="val1")
result=tf.random.truncated_normal(shape=[tens], mean=4, stddev=1)
with tf.compat.v1.Session() as val:
    new_output=val.run(result)
    print(new_output)

In the above code, we have imported the Tensorflow library and then use the tf.compat.v1.disable_eager_execution function for creating the session.

After that, we have applied the constant function for creating a tensor shape and then use the tf.random.truncated_normal() function and within this function we have assigned the shape, mean, stddev as an argument. Once you will execute this code the output displays random values from a normal distribution.

Here is the Screenshot of the following given code.

Read: TensorFlow Tensor to NumPy

TensorFlow distribution truncated_normal

• Here we are going to discuss how to use the distribution truncated_normal() function in Python TensorFlow.
• In Python, this function is a normal distribution and it is in the limit between low and high parameters while the probability density is 0 outside these limits.

Syntax:

Let’s have a look at the Syntax and understand the working of tfp.distribution.TruncatedNormal() function in Python TensorFlow.

tfp.distribution.TruncatedNormal
                                (
                                 loc,
                                 scale,
                                 low,
                                 high,
                                 validate_args=False,
                                 allow_nan_stats=True,
                                 name='TruncatedNormal'
                                )

• It consists of a few parameters
  • loc: This parameter indicates the mean of the normal distribution and the value must be a float.
  • scale: This parameter specifies the standard deviation of the normal distribution.
  • low:It is a lower limit of the normal distribution and represents floating values.
  • high: It is the upper limit of the normal distribution and indicates floating values.
  • validate_args: By default, it takes a false value and if it is true then distribution arguments are checked at run time.
  • allow_nan_stats: By default, it takes a true value and if it is ‘false’ an exception is raised if the statistics nan values are undefined.

Example:

Let’s take an example and check how to use the distribution truncated_normal() function in Python.

Source Code:

import tensorflow as tf
import tensorflow_probability as tfp
tf.compat.v1.disable_eager_execution()

new_trunc = tfp.distributions.TruncatedNormal(loc=[1., 2.], scale=2.,
                           low=[-2., 1.],
                           high=[2., 2.])
result=new_trunc.prob([0.6, 0.9])
with tf.compat.v1.Session() as val:
    new_output=val.run(result)
    print(new_output)

In the following given code, we have imported the tensorflow_probability() library with the alias name ‘tfp’.

After that, we have used the tfp.distribution.TruncatedNormal() function and within this function, we have assigned loc, scale, low, and high as an argument.

Here is the implementation of the following given code

Read: TensorFlow get shape

TensorFlow truncated_normal_initailizer

• In this Program, we will discuss how to use the truncated_normal_initializer() function in TensorFlow Python.
• In Python, the initializer means that we have to generate tensors with a normal distribution.

Syntax:

Here is the Syntax of tf.compat.v1.truncated_normal_initializer() function in Python TensorFlow.

tf.compat.v1.truncated_normal_initializer
                                         (
                                          mean=0.0,
                                          stddev=1.0,
                                          seed=None,
                                          dtype=tf.dtypes.float32
                                         )

Example:

Let’s take an example and check how to use the truncated_normal_initializer() function in Python.

Source Code:

import tensorflow as tf

new_trunc = tf.compat.v1.truncated_normal_initializer(mean=0,stddev=1,seed=4)
print(new_trunc)

You can refer to the below Screenshot.

Read: Python TensorFlow reduce_sum

TensorFlow truncated normal example

• In this example, we will discuss how to generate the random values from the truncated normal distribution in TensorFlow Python.
• To perform this particular task, we are going to use the tf.random.truncated_normal() function and this function is used to generate the random values from a truncate normal distribution.

Syntax:

Here is the Syntax of tf.random.truncated_normal() function in Python TensorFlow.

tf.random.truncated_normal
                          (
                           shape,
                           mean=0.0,
                           stddev=1.0,
                           dtype=tf.dtypes.float32,
                           seed=None,
                           name=None
                          )

Example:

import tensorflow as tf

trunc_tens=tf.random.truncated_normal([4])
print(trunc_tens)

In the above code, we have imported the Tensorflow library and then used the tf.random.truncated_normal() function and within this function, we have assigned the shape of the output tensor.

Here is the execution of the following given code.

Read: Python TensorFlow reduce_mean

TensorFlow has no attribute ‘truncated_normal’

Here we are going to discuss the error Attribute error: module ‘TensorFlow’ has no attribute ‘truncated_normal’ in Python. Basically, this error statement comes when we used the tf.truncated_normal() function.

Example:

import tensorflow as tf


tf.compat.v1.disable_eager_execution()
tens = tf.constant(10)
result=tf.truncated_normal(shape=[tens], mean=3, stddev=1)
with tf.compat.v1.Session() as val:
    new_output=val.run(result)
    print(new_output)

Here is the Screenshot of the following given code

Now let’s see the solution to this error

import tensorflow as tf

tf.compat.v1.disable_eager_execution()
tens = tf.constant(10)
result=tf.random.truncated_normal(shape=[tens], mean=3, stddev=1)
with tf.compat.v1.Session() as val:
    new_output=val.run(result)
    print(new_output)

In the above code, we have used the tf.random.truncated_normal() function instead of tf.truncated_normal() function. In Python the tf.random.truncated_normal() function is used to generate the random values from normal truncated distribution.

Here is the Output of the following given code

Read: Import error no module named TensorFlow

TensorFlow has no attribute ‘truncated_normal_initializer’

In this section, we will discuss the error Attribute error module ‘TensorFlow’ has no attribute ‘truncated_normal_initializer‘ in Python. Basically, this error statement comes when we used the tf.truncated_normal_initializer() function.

Example:

import tensorflow as tf

new_trunc = tf.truncated_normal_initializer(mean=0,stddev=1,seed=4)
print(new_trunc)

Here is the implementation of the following given code

Now let’s see the solution to this error

To get the solution to this error, you can refer to our previous topic TensorFlow truncated_normal_initailizer.

In Python TensorFlow, the latest version 2.8 has been updated most of the functions. If you are using the latest version of TensorFlow then you can apply the tf.compat.v1.truncated_normal_initializer() function to solve this error.

Also, check the following TensorFlow tutorials:

• Tensorflow iterate over tensor
• TensorFlow Graph – Detailed Guide
• TensorFlow Sparse Tensor + Examples
• Module ‘TensorFlow’ has no attribute ‘session’
• Module ‘TensorFlow’ has no attribute ‘get_default_graph’

In this Python tutorial, we have learned how to use the truncated normal function in Python. Additionally, we have covered the following topics.

• TensorFlow distribution truncated_normal
• TensorFlow truncated_normal_initailizer
• TensorFlow truncated normal example
• TensorFlow has no attribute ‘truncated_normal’
• TensorFlow has no attribute ‘truncated_normal_initializer’