Page 73 - Informatics_Practices_Fliipbook_Class12
P. 73

Dropping Columns

            In the DataFrame df, it would make sense to drop the column Month_df2. We can drop a column using the DataFrame
            method drop(), as shown below:

             >>> df.drop(['Month_df2'], axis=1)
            output:
                        Month     Rent   Utilities       Groceries   Salary   Bonus       Investments
                 0   January      1200           150           300     4000     500              1000
                 1  February      1200           170           350     4000     600               900
                 2     March      1200           160           320     4000     550              1100
                 3     April      1300           165           350     4000     600               900


                     1.  How many columns does the DataFrame df have? Hint: Note that the method drop() as used
                        above does not modify the original DataFrame df.
                     2.  On execution of the following function call, how many columns will the DataFrame df have?
                        df.drop([Month_df2], axis=1, inplace=True)

            Dropping Rows from a DataFrame

            Suppose, we wish to delete records for month 'January' and 'April', we may again use the drop the rows using the
            DataFrame method drop() by setting the column indexes to [0, 3] and keyword argument axis to 0 as shown
            below:

             >>> df.drop([0,3], axis=0)
            output:
                        Month     Rent   Utilities       Groceries    Salary    Bonus  Investments
                 1   February     1200           170            350      4000       600             900
                 2      March     1200           160            320      4000       550            1100
            Recall  that  as  the  row  axis  is  the  default  axis,  to  delete  the  rows  [0,  3],  we  could  have  also  used  the  function
            call: df.drop([0,3]).

            2.8.2 Renaming Columns of a DataFrame

            As names of the columns of DataFrame are often read input from a spreadsheet, it may be necessary to rename some
            columns to align with the data analysis task being handled.
            For example, suppose, we wish to rename the column Product as Item. The rename() method in Pandas is used
            to rename columns. It allows us to specify new names for one or more columns using a dictionary.
            The syntax for using the rename() method to rename columns is as follows:

                 df.rename(columns={'current_name': 'new_name'}, inplace=True)
             >>> groceryDF.rename(columns = {'Product': 'Item'}, inplace = True)
             >>> groceryDF.head(3)
            output:
                          Item       Category       Price      Quantity
                 0       Bread            Food         20              2
                 1        Milk            Food         60              5
                 2     Biscuit            Food         20              2

            Another way to rename columns in a Pandas DataFrame is by specifying the complete list of names of the columns,
            retaining some of the column names as it is while modifying some others, and assigning it to attribute columns of
            groceryDF, as shown below:

             >>> groceryDF.columns = ['Item Name', 'Category',            'Price',  'Quantity']
             >>> print(groceryDF.head())
                     Item Name Category  Price  Quantity
                 0       Bread     Food     20         2
                                                                             Data Handling using Pandas DataFrame  59
   68   69   70   71   72   73   74   75   76   77   78