Creating views on Pandas DataFrame
Many times while doing data analysis we are dealing with a large data set, having a lot of attributes. All the attributes are not necessarily equally important. As a result, we want to work with only a set of columns in the dataframe. For that purpose, let’s see how we can create views on the Dataframe and select only those columns that we need and leave the rest.
For link to the CSV file used in the code, click here.
Solution #1: A set of columns in the DataFrame can be selected by dropping all those columns which are not needed.
# importing pandas as pd import pandas as pd # Reading the csv file df = pd.read_csv("nba.csv") # Print the dataframe print(df) |
Output :
Now we will select all columns except the first three columns.
# drop the first three columns df.drop(df.columns[[0, 1, 2]], axis = 1) |
Output :
We can also use the names of the column to be dropped.
# drop the 'Name', 'Team' and 'Number' columns df.drop(['Name', 'Team', 'Number'], axis = 1) |
Output :
Solution #2 : We can individually select all those columns which we need and leave out the rest.
# importing pandas as pd import pandas as pd # Reading the csv file df = pd.read_csv("nba.csv") # select the first three columns # and store the result in a new dataframe df_copy = df.iloc[:, 0:3] # Print the new DataFrame df_copy |
Output :
We can also select the columns in a random manner by passing a list to the DataFrame.iloc attribute.
# select the first, third and sixth columns # and store the result in a new dataframe # The numbering of columns begins from 0 df_copy = df.iloc[:, [0, 2, 5]] # Print the new DataFrame df_copy |
Output :
Alternatively, we can also name the columns that we want to select.
# Select the below listed columns df_copy = df[['Name', 'Number', 'College']] # Print the new DataFrame df_copy |
Output :
