Index (pandas)

09 Dec 2023 in Study / Computer science on Pandas

pandas

• Choose the columns as the index of dataframe (set index)
• Initialization of index
• Rearrange index
• Sorting

Set Index

• Specific columns can be set as the index of dataframe
• Single index
  • df.set_index(“column2”) (column2 gonna be the index of this dataframe)
• Multiple index
  • df.set_index([“column1”, “column3”]) (Both column1 and column3 gonna be the index of this dataframe)
• If set_index is used, original index gonna be deleted from its dataframe

Initialization of Index

• Use “reset_index”
  • origin source changes (X) → just return
  • If we wanna change the origin dataframe, we insert “inplace=True” as a parameter
• df.reset_index()
  • Integer position index gonna be the index of dataframe
  • Original index gonna be the first column (The column’s name is “index”)

Rearrange Index

• If we wanna change the order of index, use “reindex”
  • origin source changes (X) → just return
  • “inplace = True” cannat be used
    

import pandas as pd

df = pd.DataFrame({'math' : [100, 20, 98], 'english' : [15, 75, 68], 'biology' : [98, 100, 95]}, index = ['Samoh', 'Rivera', 'Cho'])
df = df.reindex(['Cho', 'Samoh', 'Rivera'])

print(df)

• The order of index gonna be changed to ‘Cho’, ‘Samoh’, ‘Rivera’

import pandas as pd

df = pd.DataFrame({'math' : [100, 20, 98], 'english' : [15, 75, 68], 'biology' : [98, 100, 95]}, index = ['Samoh', 'Rivera', 'Cho'])
df = df.reindex(['Cho', 'Samoh', 'Rivera', 'Sanford', 'Westberry'])

print(df)

• ‘Sanford’ and ‘Westberry’ are not involved in the original dataframe : They gonna be added as new index (all of elements are NaN)
• NaN can be replaced if we insert fill_value parameter to reindex method
  • df.reindex(index_list, fill_value = 3) (If some elements in index_list are not involved in the original dataframe, they gonna be added as new index and their elements’ value is 3)

Sorting Based on Index

• Based on index, rows of dataframe can be sorted
• Use “sort_index”
  • origin source changes (X) → just return
  • If we wanna change the origin dataframe, we insert “inplace=True” as a parameter
• df.sort_index(ascending=True)
• df.sort_index(ascending=False)

import pandas as pd

df = pd.DataFrame({'math' : [100, 20, 98], 'english' : [15, 75, 68], 'biology' : [98, 100, 95]}, index = ['Samoh', 'Rivera', 'Cho'])

df = df.sort_index(ascending=True)

print(df)

• The order of index : ‘Cho’, ‘Rivera’, ‘Samoh’

Sorting Based on Value

• We can sort the dataframe by choosing a single column → It is sorted using the value of elements in this column
• Use “sort_values”
  • origin source changes (X) → just return
  • If we wanna change the origin dataframe, we insert “inplace=True” as a parameter

import pandas as pd

df = pd.DataFrame({'math' : [100, 20, 98], 'english' : [15, 75, 68], 'biology' : [98, 100, 95]}, index = ['Samoh', 'Rivera', 'Cho'])

df = df.sort_values(by='math', ascending=False)

print(df)

• By the value of elements in math, the order of index is ‘Samoh’, ‘Cho’, ‘Rivera’