Sorting the contents of a pandas DataFrame in Python

Overview:

A DataFrame is organized as a set of rows and columns identified by the row index/row labels and column index/column labels.
In Data Analysis, it is a frequent requirement to sort the DataFrame contents based on their values, either column-wise or row-wise. Also, it is a common requirement to sort a DataFrame by row index or column index.

Sorting the contents of a DataFrame by values:

The sort_values() function sorts a DataFrame by columns or by rows based on the value of its elements and returns the sortedDataFrame as a new DataFrame instance.
Either rows are sorted (axis=0) or columns are sorted (axis=1). Column ordering and row ordering is to be specified through the parameter “by”.
The parameter "ascending" controls whether the sorted DataFrame contents are in the increasing order or in the decreasing order.
The sorting method can be decided through the parameter kind – which allows one of the values among "quicksort", "mergesort", "heapsort" to be selected.

Example – Sorting the DataFrame contents by row values:

# Example Python program to sort the contents

# of a DataFrame by its values along axis=0(row-wise)

import pandas as pd

contents = [(1,3,8,9),

(17,12,5,4),

(3,13,8,19),

(3,11,9,7)

];

dataFrame = pd.DataFrame(data=contents);

dataFrame.columns = ["a","b","c","d"];

print("Contents of the DataFrame:");

print(dataFrame);

print("Contents of the DataFrame:After sorting rows based on columns a and b");

print(dataFrame.sort_values(by=["a", "b"]));

Output:

Contents of the DataFrame:

a b c d

0 1 3 8 9

1 17 12 5 4

2 3 13 8 19

3 3 11 9 7

Contents of the DataFrame:After sorting rows based on columns a and b

a b c d

0 1 3 8 9

3 3 11 9 7

2 3 13 8 19

1 17 12 5 4

Example - Sorting the DataFrame contents by column values:

# Example Python program to sort the contents

# of a DataFrame by its values along axis=1(column-wise)

import pandas as pd

contents = [("x","a","g","l"),

("o","t","o","b"),

("r","l","v","w"),

("x","a","z","d")

];

dataFrame = pd.DataFrame(data=contents, index=("r1","r2","r3","r4"));

dataFrame.columns = ["C1","C2","C3","C4"];

print("Contents of the DataFrame:");

print(dataFrame);

print("Contents of the DataFrame:After sorting columns based on rows r2 and r3");

print(dataFrame.sort_values(by=["r2","r3"], axis=1));

Output:

Contents of the DataFrame:

C1 C2 C3 C4

r1 x a g l

r2 o t o b

r3 r l v w

r4 x a z d

Contents of the DataFrame:After sorting columns based on rows r2 and r3

C4 C1 C3 C2

r1 l x g a

r2 b o o t

r3 w r v l

r4 d x z a

Sorting DataFrame contents based on DataFrame index:

Loading data into a DataFrame is typically done from external sources like a database or spreadsheet or a text file with fields separated by a delimiter( comma, pipe, tab and so on).
While loading a DataFrame from external data sources, often it is possible that the incoming data - rows or columns are not sorted based on row ids or column ids.
A DataFrame has two indexes – one for the rows and one for the columns. A DataFrame can be sorted either by its column index or row index.

Example: Sorting the DataFrame contents based on row index i.e.,Axis=0

This Python example code sorts the contents of the DataFrame based on row-index.

# Python example program to sort a pandas DataFrame

# using its row index (axis=0)

import pandas as pd

# Data as list of tuples

data = [(1, 1, 2, 3),

(1, 4, 9, 16),

(1, 5, 12, 22),

(1, 6, 15, 28),

(1, 11, 21, 1211)];

# Create a pandas DataFrame with row index

df = pd.DataFrame(data, index = (2,1,0,4,3), columns=("a","b","c","d"));

print("Contents of the DataFrame:");

print(df);

print("Contents of the DataFrame after sorting based on row index:");

print(df.sort_index(axis=0));

Output:

Contents of the DataFrame:

a b c d

2 1 1 2 3

1 1 4 9 16

0 1 5 12 22

4 1 6 15 28

3 1 11 21 1211

Contents of the DataFrame after sorting based on row index:

a b c d

0 1 5 12 22

1 1 4 9 16

2 1 1 2 3

3 1 11 21 1211

4 1 6 15 28

Example: Sorting the DataFrame contents based on column index i.e.,Axis=1

The Python example code below sorts the contents of the DataFrame based on column-index.

# Python example program to sort a pandas DataFrame

# using its column index (axis=1)

import pandas as pd

# Data as python dictionary

data = {"d":["m","n","o","p"],

"a":["a","b","c","d"],

"d":["i","j","k","l"],

"c":["e","f","g","h"]};

# Create a pandas DataFrame with column index

df = pd.DataFrame(data);

print("Contents of the DataFrame:");

print(df);

print("Contents of the DataFrame after sorting based on column index:");

print(df.sort_index(axis=1));

Output:

Contents of the DataFrame:

d a c

0 i a e

1 j b f

2 k c g

3 l d h

Contents of the DataFrame after sorting based on column index:

a c d

0 a e i

1 b f j

2 c g k

3 d h l