Comvert pandas DataFrame to numpy Array

In this tutorial, we will discuss how to convert pandas dataframe to numpy array.

Introduction

DataFrame in pandas is two dimensional data structure that will store data in two dimensional format. One dimension refers to a row and second dimension refers to a column, So It will store the data in rows and columns.

 

We create DataFrame using DataFrame() method. But this is available in pandas module, so we have to import pandas module.

Syntax:

pandas.DataFrame(data)

Where, data is the input dataframe , The data can be a dictionary that stores list of values with specified key.

 

numpy stands for numeric python which is used to perform mathematical operations on arrays.

It is a module in which we have to import from the python. 

Syntax to import:

import numpy

We can also use alias for the module 

For example,

import numpy as np

We can directly use np to call the numpy module.

Let's create a dataframe with 4 rows and 4 columns.

import pandas as pd

#create dataframe from the college data
data= pd.DataFrame({'college_id':['c-001','c-021','c-002','c-004'],

                    'college_name':["vignan university","vvit","RVR - JC","Andhra University"],

                   "college_address":["guntur","guntur","guntur","guntur"],

                    "Total Staff":['1200','3422','5644','670']

                   },index=['one','two','three','four'])

#display the dataframe
print(data)

Output:

      college_id       college_name college_address Total Staff
one        c-001  vignan university          guntur        1200
two        c-021               vvit          guntur        3422
three      c-002           RVR - JC          guntur        5644
four       c-004  Andhra University          guntur         670

Method -1 : using to_numpy()

By using this method, we can convert the entire dataframe to the numpy array.

Syntax:

dataframe.to_numpy()

where, dataframe is the input pandas dataframe.

Example:

In this example,we will convert the above dataframe to the numpy array.

import pandas as pd

#create dataframe from the college data
data= pd.DataFrame({'college_id':['c-001','c-021','c-002','c-004'],

                    'college_name':["vignan university","vvit","RVR - JC","Andhra University"],

                   "college_address":["guntur","guntur","guntur","guntur"],

                    "Total Staff":['1200','3422','5644','670']

                   },index=['one','two','three','four'])


#convert to the numpy array
print(data.to_numpy())

Output:

[['c-001' 'vignan university' 'guntur' '1200']
 ['c-021' 'vvit' 'guntur' '3422']
 ['c-002' 'RVR - JC' 'guntur' '5644']
 ['c-004' 'Andhra University' 'guntur' '670']]

If we want to get the type, use type() to get the type of the data structure.

import pandas as pd

#create dataframe from the college data
data= pd.DataFrame({'college_id':['c-001','c-021','c-002','c-004'],

                    'college_name':["vignan university","vvit","RVR - JC","Andhra University"],

                   "college_address":["guntur","guntur","guntur","guntur"],

                    "Total Staff":['1200','3422','5644','670']

                   },index=['one','two','three','four'])


#convert to the numpy array and get the type
print(type(data.to_numpy()))

Output:

<class 'numpy.ndarray'>

If we want to convert only particulr columns to numpy array, then we have to specify the column name to be converted to numpy array.

Syntax:

dataframe[['column_name']].to_numpy()

Example:

In this example we are converting college_id and college_name columns to numpy array.

import pandas as pd

#create dataframe from the college data
data= pd.DataFrame({'college_id':['c-001','c-021','c-002','c-004'],

                    'college_name':["vignan university","vvit","RVR - JC","Andhra University"],

                   "college_address":["guntur","guntur","guntur","guntur"],

                    "Total Staff":['1200','3422','5644','670']

                   },index=['one','two','three','four'])

#convert college_id column  to the numpy array
print(data[['college_id']].to_numpy())

print()

#convert  college_name column to the numpy array
print(data[['college_name']].to_numpy())

Output:

[['c-001']
 ['c-021']
 ['c-002']
 ['c-004']]

[['vignan university']
 ['vvit']
 ['RVR - JC']
 ['Andhra University']]

Method -2 : using values

By using this method, we can convert the entire dataframe to the numpy array.

Syntax:

dataframe.values

where, dataframe is the input pandas dataframe.

Example:

In this example,we will convert the above dataframe to the numpy array.

import pandas as pd

#create dataframe from the college data
data= pd.DataFrame({'college_id':['c-001','c-021','c-002','c-004'],

                    'college_name':["vignan university","vvit","RVR - JC","Andhra University"],

                   "college_address":["guntur","guntur","guntur","guntur"],

                    "Total Staff":['1200','3422','5644','670']

                   },index=['one','two','three','four'])

#convert to the numpy array
print(data.values)

print()

#convert to the numpy array and get the type
print(type(data.values))

Output:

[['c-001' 'vignan university' 'guntur' '1200']
 ['c-021' 'vvit' 'guntur' '3422']
 ['c-002' 'RVR - JC' 'guntur' '5644']
 ['c-004' 'Andhra University' 'guntur' '670']]

<class 'numpy.ndarray'>