How to create a DataFrames in Python

How to create a DataFrames in Python

Data Frame is a data structure in which data is stored in tabular form. They can also be referred as two-dimensional collection of data.

Various arithmetic operations can be performed on a particular data frame such as addition and deletion of row/column.

Data Frames can be imported from external storages and in different forms, such as SQL databases, CSV file, and Excel files.

Below are various way to create a DataFrame’s in Python:

1. Empty Data Frame
2. Data Frame using Lists
3. Data Frame using Dictionary
4. Data Frames with Index
5. Data Frames using Lists of Dictionaries
6. Data Frames using Zip() function
7. Data Frames using Dicts of Series

First requirement to create dataframe in python is Pandas library.

1. Creating an Empty Data Frame

A basic empty data frame can be created by calling the constructor called ‘Data Frame’. Consider the following example.

Example

 # importing pandas library as pd
 import pandas as pd  
 # Calling DataFrame Constructor  
 Data_frame = pd.DataFrame()  
 print(Data_frame) 

Output

 Empty Data_Frame
 Columns: []
 Index: [] 

• Creating DataFrame using Lists

DataFrame’s can also be created using list.

In the example below, we have defined a list called fruits_list which contains the names of the fruits.

Now, to make a dataframe of this list, we have called DataFrame constructor and passed fruits_list as an argument in it.

Example –

 # importing pandas as pd
 import pandas as pd 
 # string values in the list  
 fruits_list = ['Apple', 'Banana', 'Strawberry' ,'Grapes' ,'Mango' ,'Watermelon' ,'Pine Apple'] 
 # Calling DataFrame constructor on list 
 dataframe = pd.DataFrame(fruits_list) 
 print(dataframe)  

Output

                     0
 0           Apple
 1         Banana
 2   Strawberry
 3         Grapes
 4         Mango
 5 Watermelon
 6   Pine Apple 

• Creating DataFrame using Dictionary

DataFrame’s can be created using dictionaries in the same way as lists.

First step is to create a dictionary, in the example below we have created a dictionary with the name ‘dictionary’ which contains the names of the city and their pincode.

Now, to make a dataframe of this dictionary, we have called DataFrame constructor and passed ‘dictionary’ as argument in it.

Example –

 import pandas as pd 
 # Dictionary of cities and their ranking
 dictionary={'CityName':['Indore','Bhopal','Visakhapatnam','Surat','Mysore','Tiruchirappalli','New Delhi Municipal Council'], 'Rank': [1,2,3,4,5,6,7]} 
 # Creating a dataframe using DataFrame() 
 dataframe = pd.DataFrame(data) 
 # Print the output. 
 print(dataframe) 

OUTPUT

            City Name             Rank
 0                       Indore     1
 1                       Bhopal     2
 2                Visakhapatnam     3
 3                        Surat     4
 4                       Mysore     5
 5              Tiruchirappalli     6
 6  New Delhi Municipal Council     7 

• Creating DataFrame’s with Index

When any dataframe is created indexing starts from ‘0’ by default. And, if we don’t want number indexing then, we can change indexing according to our needs.

So, for changing the index’s names we pass the array of index names as an argument along with the dictionary inside the DataFrame() method.

Let’s understand with the following example:

Example –

 import pandas as pd 
 # Dictionary of cities and their ranking 
 dictionary = {'City Name': ['Indore','Bhopal','Visakhapatnam','Surat','Mysore','Tiruchirappalli','New Delhi Municipal Council'], 'Rank': [1,2,3,4,5,6,7]} 
 # Creating dataframe with indexing
 dataframe = pd.DataFrame(dictionary, index =['city1', 'city2', 'city3', 'city4','city5','city6','city7']) 
 # print the data 
 print(dataframe)  

Output

                          City Name  Rank
 city1                       Indore     1
 city2                       Bhopal     2
 city3                Visakhapatnam     3
 city4                        Surat     4
 city5                       Mysore     5
 city6              Tiruchirappalli     6
 city7  New Delhi Municipal Council     7 

• Creating DataFrame using List of Dictionaries

DataFrame’s can also be created with list of dictionaries in the same way as above.

Let’s understand the following examples:

Example 1

 import pandas as pd 
 # created list of two dicionaries
 list_of_dictionaries = [{'A': 10, 'B': 20, 'C':30}, {'A':100, 'B': 200, 'C': 300}] 
 # Creates DataFrame. 
 dataframe = pd.DataFrame(list_of_dictionaries) 
 # Print the data 
 print(dataframe)  

OUTPUT

      A    B    C
 0   10   20   30
 1  100  200  300 

Example 2

 # In this example a DataFrame is created by passing a list of dictionaries as well as indexes names
 import pandas as pd 
 # list of dictionaries is created
 list_of_dictionaries = [{'A': 2, 'B':3}, {'A': 10, 'B': 20, 'C': 30}] 
 # list of dictionaries is passed as an argument with index names and converted into DataFrame
 dataframe = pd.DataFrame(list_of_dictionaries, index =['first', 'second']) 
 # Print the dataframe 
 print(dataframe) 

 Output

          A   B     C
 first    2   3   NaN
 second  10  20  30.0 

Example 3

 # In this example DataFrame is created with the lists of dictionaries as well as with index names and coloumn names
 import pandas as pd
 # Created the lists of dictionaries
 List_of_dictionaries = [{'X': 1, 'Y': 2}, {'X': 5, 'Y': 10, 'Z': 20}]
 # converting the list of dictionaries into data frame
 dataframe1 = pd.DataFrame(List_of_dictionaries , index =['first',
                                  'second'],
                    columns =['X', 'Y'])
 # Converting the list of dictionaries into dataframe, but changing the one of the coloumn index name
 dataframe2 = pd.DataFrame(List_of_dictionaries, index =['first', 'second'], columns =['X', 'Y1'])
 # printing the first dataframe
 print (dataframe1, "\n")
 # printing the second dataframe
 print (dataframe2) 

OUTPUT

         X   Y
 first   1   2
 second  5  10 
  
         X  Y1
 first   1 NaN
 second  5 NaN 

• Creating Dataframe using Zip() Function

DataFrame’s can also be created using zip() function. The zip() function is used to merge the two lists.

In this example first two lists are merged by zip() function and new tuple is formed. Then, new tuple is converted into dataframe.

 import pandas as pd 
 # List1 
 City_Name= ['Indore','Bhopal','Visakhapatnam','Surat','Mysore','Tiruchirappalli','New Delhi Municipal Council']
 # List2 
 Rank= [1,2,3,4,5,6,7]
 # two lists are merged together using zip() function.
 # New Tuple is created after merging both the lists
 new_tuple = list(zip(City_Name, Rank)) 
 # Printing new_tuple 
 print(new_tuple) 
 # converting the new_tuple into dataframe using DataFrame()
 dataframe = pd.DataFrame(new_tuple,columns=['City Name', 'Rank']) 
 # Print data. 
 print( dataframe)  

OUTPUT

[('Indore', 1), ('Bhopal', 2), ('Visakhapatnam', 3), ('Surat', 4), ('Mysore', 5), ('Tiruchirappalli', 6), ('New Delhi Municipal Council', 7)]

                     City Name  Rank

0                       Indore     1

1                       Bhopal     2

2                Visakhapatnam     3

3                        Surat     4

4                       Mysore     5

5              Tiruchirappalli     6

6  New Delhi Municipal Council     7

• Creating DataFrame using Dicts of Series.

We can use the Dicts of series where the subsequent index is the union of all series of passed index value. Let’s understand below example.

 import pandas as pd 
 # creating the dicts of series 
 data = {
 'Data Structure' : pd.Series([60,70,86,79], index =['Pragya','Rashi','Anupriya','Prince']),
 'DBMS' : pd.Series([67,89,76,89], index =['Pragya','Rashi','Anupriya','Prince'])
 } 
 # converting data into dataframe
 dataframe = pd.DataFrame(data) 
 # print the dataframe 
 print(dataframe)  

OUTPUT

          Data Structure  DBMS

Pragya                60    67

Rashi                 70    89

Anupriya              86    76

Prince                79    89