# An Introduction to TensorFlow

# An Introduction to TensorFlow

### TensorFlow is a library that was developed by Google for solving complicated mathematical problems. This introduction to TensorFlow contains all you need to know!

Join the DZone community and get the full member experience.

Join For FreeInsight for I&O leaders on deploying AIOps platforms to enhance performance monitoring today. Read the Guide.

In this post, we are going to see some TensorFlow examples, define tensors, perform math operations using tensors, and see other machine learning examples.

## What Is TensorFlow?

TensorFlow is a library that was developed by Google for solving complicated mathematical problems, which takes a lot of time.

TensorFlow can do many things, like:

**Solve complex mathematical expressions**.**Perform machine learning techniques**in which you give it a sample of data for training, then you give another sample of data to predict the result based on the training data. This is the artificial intelligence component!**Provide GPU support**. You can use GPU (Graphical Processing Unit) instead of CPU for faster processing. There are two versions of TensorFlow: the CPU version and the GPU version.

Before we start working with TensorFlow examples, we need to know some basics.

## What Is a Tensor?

A tensor is the main blocks of data that TensorFlow uses. They're like the variables that TensorFlow uses to work with data. Each tensor has a dimension and a type.

The dimension refers to the rows and columns of the tensor. You can define one-dimensional tensors, two-dimensional tensors, and three-dimensional tensors, as we will see later.

The type refers to the data type for the elements of the tensor.

## Define One-Dimensional Tensor

To define a tensor, we will create a NumPy array or a Python list and convert it to a tensor using the `tf_convert_to_tensor`

function.

We will use NumPy to create an array like this:

```
import numpy as np
arr = np.array([1, 5.5, 3, 15, 20])
```

The results show the dimension and the shape of the array.

```
import numpy as np
arr = np.array([1, 5.5, 3, 15, 20])
print(arr)
print (arr.ndim)
print (arr.shape)
print (arr.dtype)
```

It looks like the Python list but here there is no comma between the items.

Now we will convert this array to a tensor using the `tf_convert_to_tensor`

function.

```
import numpy as np
import tensorflow as tf
arr = np.array([1, 5.5, 3, 15, 20])
tensor = tf.convert_to_tensor(arr,tf.float64)
print(tensor)
```

From the results, you can see the tensor definition, but you can’t see the tensor elements.

To see the tensor elements, you can run a session like this:

```
import numpy as np
import tensorflow as tf
arr = np.array([1, 5.5, 3, 15, 20])
tensor = tf.convert_to_tensor(arr,tf.float64)
sess = tf.Session()
print(sess.run(tensor))
print(sess.run(tensor[1]))
```

## Define Two-Dimensional Tensor

This is done the same way as the one-dimensional array, but this time, we will define the array like this:

`arr = np.array([(1, 5.5, 3, 15, 20),(10, 20, 30, 40, 50), (60, 70, 80, 90,100)])`

And you can convert it to a tensor like this:

```
import numpy as np
import tensorflow as tf
arr = np.array([(1, 5.5, 3, 15, 20),(10, 20, 30, 40, 50), (60, 70, 80, 90, 100)])
tensor = tf.convert_to_tensor(arr)
sess = tf.Session()
print(sess.run(tensor))
```

Now you know how to define tensors. What about performing some math operations between them?

## Performing Math on Tensors

Suppose that we have two arrays like this:

```
arr1 = np.array([(1,2,3),(4,5,6)])
arr2 = np.array([(7,8,9),(10,11,12)])
```

We need to get the sum of them. You can perform many math operations using TensorFlow.

You can use the add function like this:

`arr3 = tf.add(arr1,arr2)`

So the whole code will be like this:

```
import numpy as np
import tensorflow as tf
arr1 = np.array([(1,2,3),(4,5,6)])
arr2 = np.array([(7,8,9),(10,11,12)])
arr3 = tf.add(arr1,arr2)
sess = tf.Session()
tensor = sess.run(arr3)
print(tensor)
```

You can multiply arrays like this:

```
import numpy as np
import tensorflow as tf
arr1 = np.array([(1,2,3),(4,5,6)])
arr2 = np.array([(7,8,9),(10,11,12)])
arr3 = tf.multiply(arr1,arr2)
sess = tf.Session()
tensor = sess.run(arr3)
print(tensor)
```

Now you got the idea.

## Three-Dimensional Tensor

We've seen how to work with one and two-dimensional tensors. Now, we will see the three-dimensional tensors. But this time, we won’t use numbers; we will use an RGB image where each piece of the image is specified by x, y, and z coordinates.

These coordinates are the width, height, and color depth.

First, let’s import the image using `matplotlib`

. You can install `matplotlib`

using `pip`

if it’s not installed on your system.

Now, put your file in the same directory as your Python file and import the image using `matplotlib`

like this:

```
import matplotlib.image as img
myfile = "likegeeks.png"
myimage = img.imread(myfile)
print(myimage.ndim)
print(myimage.shape)
```

As you can see, it’s a three-dimensional image where the width is 150, the height is 150, and the color depth is 3.

You can view the image like this:

```
import matplotlib.image as img
import matplotlib.pyplot as plot
myfile = "likegeeks.png"
myimage = img.imread(myfile)
plot.imshow(myimage)
plot.show()
```

Cool!

What about manipulating the image using TensorFlow? Super easy.

## Crop or Slice Image Using TensorFlow

First, we put the values on a placeholder like this:

`myimage = tf.placeholder("int32",[None,None,3])`

To slice the image, we will use the slice operator like this:

`cropped = tf.slice(myimage,[10,0,0],[16,-1,-1])`

Finally, run the session:

`result = sess.run(cropped, feed_dict={slice: myimage})`

Then you can see the result image using `matplotlib`

.

So the whole code will be like this:

```
import tensorflow as tf
import matplotlib.image as img
import matplotlib.pyplot as plot
myfile = "likegeeks.png"
myimage = img.imread(myfile)
slice = tf.placeholder("int32",[None,None,3])
cropped = tf.slice(myimage,[10,0,0],[16,-1,-1])
sess = tf.Session()
result = sess.run(cropped, feed_dict={slice: myimage})
plot.imshow(result)
plot.show()
```

Awesome!

## Transpose Images Using Tensorflow

In this TensorFlow example, we will do a simple transformation using TensorFlow.

First, specify the input image and initialize TensorFlow variables:

```
myfile = "likegeeks.png"
myimage = img.imread(myfile)
image = tf.Variable(myimage,name='image')
vars = tf.global_variables_initializer()
```

Then we will use the `transpose`

function, which flips the 0 and 1 axes of the input grid:

```
sess = tf.Session()
flipped = tf.transpose(image, perm=[1,0,2])
sess.run(vars)
result=sess.run(flipped)
```

Then, you can show the resulting image using `matplotlib`

.

```
import tensorflow as tf
import matplotlib.image as img
import matplotlib.pyplot as plot
myfile = "likegeeks.png"
myimage = img.imread(myfile)
image = tf.Variable(myimage,name='image')
vars = tf.global_variables_initializer()
sess = tf.Session()
flipped = tf.transpose(image, perm=[1,0,2])
sess.run(vars)
result=sess.run(flipped)
plot.imshow(result)
plot.show()
```

All of these TensorFlow examples show you how easy it os to work with TensorFlow.

I hope you find the post useful. Keep coming back!

TrueSight is an AIOps platform, powered by machine learning and analytics, that elevates IT operations to address multi-cloud complexity and the speed of digital transformation.

Published at DZone with permission of Seco Max , DZone MVB. See the original article here.

Opinions expressed by DZone contributors are their own.

## {{ parent.title || parent.header.title}}

## {{ parent.tldr }}

## {{ parent.linkDescription }}

{{ parent.urlSource.name }}