In this Python tutorial, we will learn** how to use TensorFlow reduce_mean() in Python**. Also, we will cover the following topics.

- TensorFlow reduce_mean with mask
- TensorFlow reduce_mean nan
- TensorFlow reduce mean squared error
- TensorFlow reduce mean ignore zero
- TensorFlow reduce_mean numpy
- TensorFlow reduce_mean reduction_indices
- TensorFlow reduce mean ignore nan
- TensorFlow reduce_mean keepdims

## Python TensorFlow reduce_mean

- In this section, we will learn how to use the
**tf.reduce_mean()**function in TensorFlow Python. - To perform this particular task, we are going to use the
**tf.math.reduce_mean()**function and this function is available in TensorFlow version**2.x**. - In Python TensorFlow, the
**tf.math.reduce_mean()**function is used to calculate the mean of values across dimensions of an input tensor.

**Syntax:**

Let’s have a look at the syntax and understand the working of **tf.math.reduce_mean() **function.

```
tf.math.reduce_mean
(
input_tensor,
axis=None,
keepdims=False,
name=None
)
```

- It consists of a few parameters
**input_tensor:**This parameter indicates the tensor which we want to reduce and it should always be a numeric type.**axis:**By default, it takes none value, and once you will use this in function then by default all dimensions will reduce.**keepdims:**This parameter will check the condition if it is true then it reduces the length of the rank tensor. By default, it takes False value.**name:**It is an optional parameter and it indicates the name for the operation.

**Example:**

Let’s take an example and check how to get the mean value of the input tensor.

```
import tensorflow as tf
new_tensor = tf.constant([[13,69,55],
[245,78,94]])
new_result = tf.math.reduce_mean(new_tensor,1)
print("Sum of nan values:",new_result)
```

In the above code we have imported the TensorFlow library and then created a tensor by using the **tf.constant()** function. After that, we have used the** tf.math.reduce_mean() **function and within this function, we have assigned the tensor and **axis=1 **as an argument.

Here is the Screenshot of the following given code.

As you can see in the Screenshot the output displays the mean value of the tensor.

Also, check: TensorFlow Tensor to numpy

## TensorFlow reduce_mean with mask

- In this section, we will discuss how to use the mast in
**reduce_mean()**function. - To do this task, we are going to use the
**tf.boolean_mask()**function and it is used to compute boolean mask to tensor this method is available in the TensorFlow package. - In this method, the mask shape must match the first K dimension of the tensor’s shape.

**Syntax:**

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

```
tf.boolean_mask
(
tensor,
mask,
axis=None,
name='boolean_mask',
)
```

- It consists of a few parameters
**tensor:**This parameter indicates the input tensor and it can be a**n-dimensional**tensor.**mask:**It is a**K-D**boolean Tensor and it will set the condition**K<=N**, where K must be represented as statically.**axis:**By default it takes none value and it specifies the axis in tensor to mask.

**Example:**

```
import tensorflow as tf
import numpy as np
tens=[14,67,89,25]
new_mask=np.array([True,False,False,True])
result=tf.boolean_mask(tens, new_mask)
print("mask value:",result)
new_output=tf.math.reduce_mean(result)
print("Mean value of mask:",new_output)
```

- In the above code, we have created a mask by using the
**np.array()**function and assigned boolean values. After that, we have used the**tf.boolean_mask()**function and within this function, we have passed the tensor and mask as an argument. - Once you will execute this function the output displays the True value along with associated tensor values.
- After that, we have used the
**tf.math.reduce_mean()**function and pass**‘result’**as an argument and it will return the mean value of the mask.

Here is the execution of the following given code

Read: TensorFlow get shape

## TensorFlow reduce_mean nan

- Here we are going to discuss how to reduce nan values by using the reduce_mean() function in TensorFlow Python.
- In this example, we are going to get the mean value of nan values. To do this task first we will import the numpy library for
**np.nan**values and then we are going to create the tensor by using the**tf.constant()**function and inside this function, we have set the nan values. In Python, the nan stands for**‘not a number’**. - Next, we will declare a variable
**‘new_result’**and assign the**tf.math.reduce_mean()**function and within this function we are going to set the input tensor and axis as an argument.

**Example:**

```
import tensorflow as tf
import numpy as np
new_tensor = tf.constant([[np.nan,np.nan,np.nan,15,76],
[np.nan,np.nan,np.nan,24,89]])
new_result = tf.math.reduce_mean(new_tensor,[0,1])
print("Sum of nan values:",new_result)
```

Here is the execution of the following given code

As you can see in the Screenshot the output displays the mean of nan values.

Read: Python TensorFlow random uniform

## TensorFlow reduce mean squared error

- In this section, we will discuss how to use the mean squared error function in TensorFlow Python.
- To perform this particular task we are going to use the
**tf.compat.v1.losses.mean_squared_error()**function and this function is available in Tensorflow packages.

**Syntax:**

Here is the Syntax of** tf.compat.v1.losses.mean_squared_error()** function

```
tf.compat.v1.losses.mean_squared_error
(
labels,
prediction,
weights=1.0,
scope=None,
)
```

- It consists of a few parameters
**labels:**This parameter indicates the parameter inside the function.**prediction:**The prediction parameter is used in**call()**method.

**Example:**

```
import tensorflow as tf
y_true = tf.constant([[4.6, 7.3, 3.2],
[4.1,5.8,7.2]])
y_pred = tf.constant([[2.4, 4.6, 9.7],
[1.2,2.3,1.6]])
result=tf.compat.v1.losses.mean_squared_error(y_true,y_pred)
print("Reduce mean squared error:",result)
```

In the following given code, we have created a tensor by using the **tf.constant()** function and then we have used the **tf.compat.v1.losses.mean_squared_error()** function and within this function, we assigned the labels and prediction as an argument.

Here is the implementation of the following given code.

Read: Module ‘TensorFlow’ has no attribute ‘session’

## TensorFlow reduce mean ignore zero

- In this section, we will learn how to ignore zero values in a tensor by using the
**reduce_mean()**function in Python. - To do this task we are going to use the
**tf.cast()**function and this function is used to caste an input tensor to a new type within this function we will set the condition**new_tensor!=0**along with datatype. - Next, we will use the
**tf.reduce_sum()**function and divide it with tensor and it will ignore zero value from the tensor.

**Syntax:**

Here is the Syntax of **tf.cast()** function.

```
tf.cast
(
x,
dtype,
name=None
)
```

- It consists of a few parameters
**x:**This parameter indicates the input tensor.**name:**This is an optional parameter and by default it takes none value.

**Example:**

```
import tensorflow as tf
new_tensor= tf.constant([[2,0,3],[12,34,0],[0,23,31]])
m = tf.cast(new_tensor != 0,tf.int32)
z = tf.reduce_sum(new_tensor, -1) / tf.reduce_sum(m, -1)
print(z)
```

Here is the Screenshot of the following given code

Read: Import error no module named TensorFlow

## TensorFlow reduce_mean numpy

- In this Program, we will learn how to work numpy compatibility in
**tf.reduce_mean()**function. - To perform this particular task, we are going to use the
**tf.reduce_mean()**function and within this function, we have assigned the tensor as an argument. - And it is used to specify the output type and by default, it takes
**float32**.

**Example:**

```
import tensorflow as tf
tensor = tf.constant([12, 13.5, 1, 1.5])
tf.reduce_mean(tensor)
```

Here is the execution of the following given code.

As you can see in the Screenshot the output displays the mean value in NumPy.

Read: Python TensorFlow one_hot

## TensorFlow reduce_mean reduction_indices

- In this section, we will discuss how to use the reduction_indices parameter in the
**tf.reduce_mean()**function. - To perform this particular task, we are going to use the
**tf.compat.v1.reduce_mean()**function and this method will help the user to calculate the mean of values across dimensions of an input tensor.

**Syntax:**

Let’s have a look at the Syntax and understand the working of **tf.compat.v1.reduce_mean()** function.

```
tf.compat.v1.reduce_mean
(
input_tensor,
axis=None,
keepdims=None,
name=None,
reduction_indices=None,
keep_dims=None
)
```

**Example:**

```
import tensorflow as tf
tensor = tf.constant([14, 15.5, 12, 17.5])
z=tf.compat.v1.reduce_mean(tensor, reduction_indices=[0])
print(z)
```

In the above code we have imported the TensorFlow library and then created the tensor by using the **tf.constant()** function.

After that we have used the** tf.compat.v1.reduce_mean()** function and within this function we have used the** reduction_indices[0]** parameter.

You can refer to the below Screenshot

Read: Python TensorFlow Placeholder

## TensorFlow reduce mean ignore nan

- In this section, we will discuss how to ignore nan values by using the
**reduce_mean()**function in Python TensorFlow. - To do this task we are going to use the tf.where function along with
**tf.math.is_nan()**and this function will help the user to remove the nan values from the tensor. - After removing the tensor we are going to use the
**tf.reduce_mean()**function to calculate the mean values of the updated tensor.

**Syntax:**

Here is the Syntax of** tf.math.is_nan()** function

```
tf.math.is_nan
(
x,
name=None,
)
```

**Example:**

Let’s take an example and understand the working **tf.math.is_nan()** function.

**Source Code:**

```
import tensorflow as tf
import numpy as np
tens1 = tf.constant([[16, 67], [24, np.nan]])
tens2 = tf.constant([[np.nan, np.nan], [np.nan, 56]])
c = tf.where(tf.math.is_nan(tens1), tens2, tens1)
result=tf.reduce_mean(c)
print(result)
```

Here is the execution of the following given code

Read: Tensorflow custom loss function

## TensorFlow reduce_mean keepdims

- In this Program, we will discuss how to use the keepdims parameter in
**reduce_mean()**function. - In this function, the keepdims parameter will check the condition if the value is true the rank of the input tensor is reduced by
**1**, and by default, it takes a false value.

**Syntax:**

Here is the Syntax of **tf.math.reduce_mean() **function

```
tf.math.reduce_mean
(
input_tensor,
axis=None,
keepdims=False,
name=None
)
```

**Example:**

```
import tensorflow as tf
new_tensor = tf.constant([[67,145,267],
[18,29,31]])
new_result = tf.math.reduce_mean(new_tensor,1,keepdims=False)
print("Sum of nan values with keepdims axis:",new_result)
```

In the following given code first, we have imported the TensorFlow library and then for creating a tensor we have used the** tf.constant()** function. After that, we have used the **tf.math.reduce_mean()** function and inside the function we have set the tensor, axis, and **keepdims=False **as an argument.

Here is the implementation of the following given code

YOu may also like to read the following Python TensorFlow tutorials.

So, in this Python tutorial, we have learned** how to use TensorFlow reduce_mean() in Python**. Also, we have covered the following topics.

- TensorFlow reduce_mean pytorch
- TensorFlow reduce_mean with mask
- TensorFlow reduce_mean nan
- TensorFlow reduce mean squared error
- TensorFlow reduce mean layer
- TensorFlow reduce mean ignore zero
- TensorFlow reduce_mean numpy
- TensorFlow reduce_mean reduction_indices
- TensorFlow reduce mean ignore nan
- TensorFlow reduce_mean keepdims

I am Bijay Kumar, a Microsoft MVP in SharePoint. Apart from SharePoint, I started working on Python, Machine learning, and artificial intelligence for the last 5 years. During this time I got expertise in various Python libraries also like Tkinter, Pandas, NumPy, Turtle, Django, Matplotlib, Tensorflow, Scipy, Scikit-Learn, etc… for various clients in the United States, Canada, the United Kingdom, Australia, New Zealand, etc. Check out my profile.