With the help of Pandas, we can perform many functions on data set like Slicing, Indexing, Manipulating, and Cleaning Data frame.
Case 1: Slicing Pandas Data frame using DataFrame.iloc[]
Example 1: Slicing Rows
Python3
# importing pandas libraryimport pandas as pd# Initializing the nested list with Data setplayer_list = [['M.S.Dhoni', 36, 75, 5428000], ['A.B.D Villers', 38, 74, 3428000], ['V.Kholi', 31, 70, 8428000], ['S.Smith', 34, 80, 4428000], ['C.Gayle', 40, 100, 4528000], ['J.Root', 33, 72, 7028000], ['K.Peterson', 42, 85, 2528000]]# creating a pandas dataframedf = pd.DataFrame(player_list, columns=['Name', 'Age', 'Weight', 'Salary'])# data frame before slicingdf |
Output:
Python3
# Slicing rows in data framedf1 = df.iloc[0:4]# data frame after slicingdf1 |
Output:
In the above example, we sliced the rows from the data frame.
Example 2: Slicing Columns
Python3
# importing pandas libraryimport pandas as pd# Initializing the nested list with Data setplayer_list = [['M.S.Dhoni', 36, 75, 5428000], ['A.B.D Villers', 38, 74, 3428000], ['V.Kholi', 31, 70, 8428000], ['S.Smith', 34, 80, 4428000], ['C.Gayle', 40, 100, 4528000], ['J.Root', 33, 72, 7028000], ['K.Peterson', 42, 85, 2528000]]# creating a pandas dataframedf = pd.DataFrame(player_list, columns=['Name', 'Age', 'Weight', 'Salary'])# data frame before slicingdf |
Output:
Python3
# Slicing columnss in data framedf1 = df.iloc[:, 0:2]# data frame after slicingdf1 |
Output:
In the above example, we sliced the columns from the data frame.
Case 2: Indexing Pandas Data frame
Python3
# importing pandas libraryimport pandas as pd# Initializing the nested list with Data setplayer_list = [['M.S.Dhoni', 36, 75, 5428000], ['A.B.D Villers', 38, 74, 3428000], ['V.Kholi', 31, 70, 8428000], ['S.Smith', 34, 80, 4428000], ['C.Gayle', 40, 100, 4528000], ['J.Root', 33, 72, 7028000], ['K.Peterson', 42, 85, 2528000]]# creating a pandas dataframe and indexing it using Aplhabetsdf = pd.DataFrame(player_list, columns=['Name', 'Age', 'Weight', 'Salary'], index=['A', 'B', 'C', 'D', 'E', 'F', 'G'])# Displaying data framedf |
Output:
In the above example, we do indexing of the data frame.
Case 3: Manipulating Pandas Data frame
Manipulation of the data frame can be done in multiple ways like applying functions, changing a data type of columns, splitting, adding rows and columns to a data frame, etc.
Example 1: Applying lambda function to a column using Dataframe.assign()
Python3
# importing pandas libraryimport pandas as pd# creating and initializing a listvalues = [['Rohan', 455], ['Elvish', 250], ['Deepak', 495], ['Sai', 400], ['Radha', 350], ['Vansh', 450]]# creating a pandas dataframedf = pd.DataFrame(values, columns=['Name', 'Univ_Marks'])# Applying lambda function to find percentage of# 'Univ_Marks' column using df.assign()df = df.assign(Percentage=lambda x: (x['Univ_Marks'] / 500 * 100))# displaying the data framedf |
Output:
In the above example, the lambda function is applied to the ‘Univ_Marks’ column and a new column ‘Percentage’ is formed with the help of it.
Another Method:
Another way of obtaining the above output is by creating a new column by using the [ ] brackets with the new column name at the left side of the assignment and operation to be performed on right side of the assignment.
Note: This method is used to create a new column derived from existing column(s) of the data frame.
Python3
# importing pandas libraryimport pandas as pd# creating and initializing a listvalues = [['Rohan', 455], ['Elvish', 250], ['Deepak', 495], ['Sai', 400], ['Radha', 350], ['Vansh', 450]]# creating a pandas dataframedf = pd.DataFrame(values, columns=['Name', 'Univ_Marks'])# displaying the data frameprint('Data frame before calculating percentage\n')print(df)print('\nData frame with Percentage Column\n')# Creating new column Percentage derived from Univ_Marksdf["Percentage"] = df["Univ_Marks"]/500*100# displaying the data frameprint(df) |
Output:
Example 2: Sorting the Data frame in Ascending order
Python3
# importing pandas libraryimport pandas as pd# Initializing the nested list with Data setplayer_list = [['M.S.Dhoni', 36, 75, 5428000], ['A.B.D Villers', 38, 74, 3428000], ['V.Kholi', 31, 70, 8428000], ['S.Smith', 34, 80, 4428000], ['C.Gayle', 40, 100, 4528000], ['J.Root', 33, 72, 7028000], ['K.Peterson', 42, 85, 2528000]]# creating a pandas dataframedf = pd.DataFrame(player_list, columns=['Name', 'Age', 'Weight', 'Salary'])# Sorting by column 'Weight'df.sort_values(by=['Weight']) |
Output:
In the above example, we sort the data frame by column ‘Weight”.
Case 4: Cleaning Pandas Data frame
Python3
# importing pandas and Numpy librariesimport pandas as pdimport numpy as np# Initializing the nested list with Data setplayer_list = [['M.S.Dhoni', 36, 75, 5428000], ['A.B.D Villers', np.nan, 74, np.nan], ['V.Kholi', 31, 70, 8428000], ['S.Smith', 34, 80, 4428000], ['C.Gayle', np.nan, 100, np.nan], [np.nan, 33, np.nan, 7028000], ['K.Peterson', 42, 85, 2528000]]# creating a pandas dataframedf = pd.DataFrame(player_list, columns=['Name', 'Age', 'Weight', 'Salary'])df |
Output:
Python3
# Checking for missing valuesdf.isnull().sum() |
Output:
Python3
# dropping or cleaning the missing datadf = df.dropna()df |
Output:
In the above example, we clean all the missing values from the data set.
