How to rank a Pandas DataFrame?
How to rank a Pandas DataFrame?
Recipe Objective
While working on a dataset we sometimes need to get ranks of the columns based on the values in other features, rank can be defined in many ways like based on ascending order or decending order of the values in the feature.
This python source code does the following :
1. Creates and converts data dictionary into pandas dataframe
2. Creates new columns in the dataframe
3. Ranks dataframe in ascending and descending order
So this is the recipe on how we rank a Pandas DataFrame.
Step 1 - Import the library
import pandas as pd
We have only imported pandas which is needed.
Step 2 - Setting up the Data
We have created a dictionary of data and passed it in pd.DataFrame to make a dataframe with columns 'first_name', 'last_name', 'age', 'Comedy_Score' and 'Rating_Score'.
raw_data = {'first_name': ['Sheldon', 'Raj', 'Leonard', 'Howard', 'Amy'],
'last_name': ['Copper', 'Koothrappali', 'Hofstadter', 'Wolowitz', 'Fowler'],
'age': [42, 38, 36, 41, 35],
'Comedy_Score': [9, 7, 8, 8, 5],
'Rating_Score': [25, 25, 49, 62, 70]}
df = pd.DataFrame(raw_data, columns = ['first_name', 'last_name', 'age',
'Comedy_Score', 'Rating_Score'])
print(df)
Step 3 - Ranking the dataframe
We want to rank the dataframe on the basis of column 'age', for better understanding we will rank on ascending as well as decending order of age. But before using rank function let us first look into its parameters.
- axis : It is bool in which 0 signifies rows and 1 signifies column and by default it is 0.
- method : In this we have to pass the method of ranking the dataframe, it can be 'average', 'min', 'max', 'first' and 'dense'. By default it is set to average.
- na_option : This to decide if we want to rank NaN values as NaN or we have give the higest or lowest rank to it. By default it is set to keep.
- ascending : This is a bool feature in which we have to especify that we want the ranking as ascending or decending.
df['Hierarchy_Rank'] = df['age'].rank(ascending=True)
print(df)
df['Hierarchy_Rank'] = df['age'].rank(ascending=False)
print(df)
So the output comes as
first_name last_name age Comedy_Score Rating_Score 0 Sheldon Copper 42 9 25 1 Raj Koothrappali 38 7 25 2 Leonard Hofstadter 36 8 49 3 Howard Wolowitz 41 8 62 4 Amy Fowler 35 5 70 first_name last_name age Comedy_Score Rating_Score Hierarchy_Rank 0 Sheldon Copper 42 9 25 5.0 1 Raj Koothrappali 38 7 25 3.0 2 Leonard Hofstadter 36 8 49 2.0 3 Howard Wolowitz 41 8 62 4.0 4 Amy Fowler 35 5 70 1.0 first_name last_name age Comedy_Score Rating_Score Hierarchy_Rank 0 Sheldon Copper 42 9 25 1.0 1 Raj Koothrappali 38 7 25 3.0 2 Leonard Hofstadter 36 8 49 4.0 3 Howard Wolowitz 41 8 62 2.0 4 Amy Fowler 35 5 70 5.0
