Pandas nlargest(), nsmallest(), swap_level(), stack(), unstack(), swap_axes()

The Pandas DataFrame/Series has several methods to re-shape, sort, and transpose the data. When applied to a DataFrame/Series, these methods evaluate and modify the data to accommodate the selections.

---

Preparation

Before any data manipulation can occur, two (2) new libraries will require installation.

• The Pandas library enables access to/from a DataFrame.
• The NumPy library supports multi-dimensional arrays and matrices in addition to a collection of mathematical functions.

To install these libraries, navigate to an IDE terminal. At the command prompt ($), execute the code below. For the terminal used in this example, the command prompt is a dollar sign ($). Your terminal prompt may be different.

$ pip install pandas

Hit the <Enter> key on the keyboard to start the installation process.

$ pip install numpy

Hit the <Enter> key on the keyboard to start the installation process.

If the installations were successful, a message displays in the terminal indicating the same.

---

Feel free to view the PyCharm installation guide for the required libraries.

• How to Install Pandas on PyCharm
• How to Install NumPy on PyCharm

---

Add the following code to the top of each code snippet. This snippet will allow the code in this article to run error-free.

import pandas as pd
import numpy as np 

---

DataFrame nlargest()

The nlargest() method retrieves and returns the first n (number) of rows containing the largest column values in descending order. This method is similar to df.sort_values().

Click here for additional details.

Filtering Pandas Dataframe with DF.QUERY, .Nlargest and .Nsmallest | LEARNEREA

The syntax for this method is as follows:

DataFrame.nlargest(n, columns, keep='first')

Parameter	Argument
n	This parameter is an integer that returns the specified (n) rows from a DataFrame/Series.
columns	This parameter is a list or list of labels to order the DataFrame/Series.
keep	This parameter deals with how to handle duplicate values. The options are 'first', 'last' and 'all'. – first/last organize the first/last occurrences. – all: includes duplicates.

For this example, the countries.csv file containing nine (9) rows reads in. The top three (3) countries with the highest areas display in descending order.

💡 Note: Click here to download this file. Move this file to the current working directory.

Code – Example 1

df = pd.read_csv('countries.csv')
result = df.nlargest(3, 'Area')
print(result)

• Line [1] reads in a comma-delimited CSV file and saves to a DataFrame (df).
• Line [2] determines the top three (3) countries with the highest area and sorts in descending order. The output saves to result.
• Line [3] outputs the result to the terminal.

Output

result

	Country	Capital	Population	Area
5	Russia	Moscow	146748590	17098246
6	USA	Washington	328239523	9833520
7	China	Beijing	1400050000	9596961

For this example, the countries.csv file containing nine (9) rows reads in. The top three (3) countries with the highest areas display in descending order. In this example, the numeric fields are formatted using a lambda!

Code – Example 2

df = pd.read_csv('countries.csv')
df = df.nlargest(3, 'Population')

df['Area'] = df['Area'].apply(lambda x: '{:,}'.format(x))
df['Population'] = df['Population'].apply(lambda x: '{:,}'.format(x))
print(df)

• Line [1] reads in a comma-delimited CSV file and saves to a DataFrame (df).
• Line [2] determines the top three (3) countries with the highest population and sorts in descending order. The output saves to df.
• Line [3-4] format the Population and Area columns with commas. The df updates accordingly.
• Line [5] outputs the DataFrame to the terminal.

Output

df

	Country	Capital	Population	Area
7	China	Beijing	1,400,050,000	9,596,961
8	India	Dheli	1,352,642,280	3,287,263
6	USA	Washington	328,239,523	9,833,520

---

DataFrame nsmallest()

The nsmallest() method retrieves and returns the first n (number) of rows that contain the smallest column values in ascending order.

Filtering Pandas Dataframe with DF.QUERY, .Nlargest and .Nsmallest | LEARNEREA

The syntax for this method is as follows:

DataFrame.nsmallest(n, columns, keep='first')

Parameter	Argument
n	This parameter is an integer that returns the specified (n) rows from a DataFrame/Series.
columns	This parameter is a list or list of labels to order the DataFrame/Series.
keep	This parameter deals with how to handle duplicate values. The options are 'first', 'last' and 'all'. – first/last organize the first/last occurrences. – all: includes duplicates.

N:	This parameter is an integer and returns the specified (n) rows from a DataFrame/Series.
Columns:	This parameter is a list or list of labels to order the DataFrame/Series.
Keep:	This parameter deals with how to handle duplicate values. The options are ‘first’, ‘last’ and ‘all’. – first/last organize the first/last occurrences. – all: includes duplicates.

For this example, the finxters.csv file containing fictitious Finxter user information reads in. The users with the smallest number of puzzles solved returns in ascending order. In this example, the numeric fields are formatted using a lambda.

df = pd.read_csv('finxters.csv')
df = df.nsmallest(3, 'Solved')

df['Solved'] = df['Solved'].apply(lambda x: '{:,}'.format(int(x)))
df['Incorrect'] = df['Incorrect'].apply(lambda x: '{:,}'.format(int(x)))
df['Recurring'] = df['Recurring'].apply(lambda x: '${:,}'.format(x))
print(df)

• Line [1] reads in a comma-delimited CSV file and saves to a DataFrame (df).
• Line [2] determines the three (3) users with the smallest number of puzzles solved and sorts in ascending order. The output saves to df.
• Line [3-4] converts to an integer and formats the Solved and Incorrect columns with commas. The df updates accordingly.
• Line [5] formats the Recurring column with a dollar sign and a comma. The df updates accordingly.
• Line [6] outputs the DataFrame to the terminal.

Output

	FID	Start	First_Name	Last_Name	…	Solved	Incorrect	Recurring	Taxes
2	30022331	11/1/2021	Peter	Dunn	…	15	9	9.98	15
18	3002285	16/6/2021	Jack	Thompson	…	91	18	15.98	18
42	30024622	6/10/2021	Jan	Martin	…	995	37	9.98	10

---

DataFrame swap_level()

The swaplevel() method trades the contents of level i with the contents of level j in a MultiIndex DataFrame/Series.

The syntax for this method is as follows:

DataFrame.swaplevel(i=- 2, j=- 1, axis=0)

Parameter	Description
i, j	These parameters can be an integer/string. They are the indexes to be swapped.
axis	If zero (0) or index is selected, apply to each column. Default is 0 (column). If zero (1) or columns, apply to each row.

For this example, there are five (5) students. Each student has some associated data with it. Grades generate by using np.random.randint() from the NumPy library.

index = [(1001, 'Micah Smith', 14), (1001, 'Philip Jones', 15), 
         (1002, 'Ben Grimes', 16), (1002, 'Alicia Heath', 17), 
         (1002, 'Arch Nelson', 18)]
m_index = pd.MultiIndex.from_tuples(index)
grades_lst = np.random.randint(45,100,size=5)
df = pd.DataFrame({"Grades": grades_lst}, index=m_index)
print(df)

result = df.swaplevel(0,1)
print(result)

• Line [1] creates a list of tuples. Each tuple contains three (3) values. The output saves to the index.
• Line [2] creates a MultiIndex from the List of Tuples created on line [1] and saves it to m_index.
• Line [3] generates five (5) random grades between the specified range and saves them to grades_lst.
• Line [4] creates a DataFrame from the variables created on lines [1-3] and saves to df.
• Line [5] outputs the DataFrame to the terminal.
• Line [6] swaps out the levels as specified. The output saves to result.
• Line [7] outputs the result to the terminal.

Output

df

			Grades
1001	Micah Smith	14	55
	Philip Jones	15	74
1002	Ben Grimes	16	93
	Alicia Heath	17	93
	Arch Nelson	18	63

result

			Grades
Micah Smith	1001	14	55
Philip Jones	1001	15	74
Ben Grimes	1002	16	93
Alicia Heath	1002	17	93
Arch Nelson	1002	18	63

---

DataFrame stack()

The stack() method returns a re-shaped Multi-Level index DataFrame/Series containing at minimum one (1) or more inner levels. A pivot occurs on the new levels using the columns of the DataFrame/Series.

#6 Pandas: Stack Unstack | Pandas Tutorial | Pandas dataframes: Reshaping, Stacking Unstacking

💡 Note: If a single level, the output returns as a Series. If multi-level, the new level(s) are retrieved from the said levels and return a DataFrame.

The syntax for this method is as follows:

DataFrame.stack(level=- 1, dropna=True)

level	This parameter is the level(s) to stack on the selected axis. Levels can be a string, integer, or list. By default, -1 (last level).
dropna	This parameter determines if rows containing missing values drop. True, by default.

We have two (2) students with relevant details that save to a DataFrame. The code below displays the original DataFrame and the DataFrame using the stack() method.

df = pd.DataFrame([[8, 7], [7, 5]],
                  index=['Micah', 'Philip'],
                  columns=['Age', 'Grade'])
print(df)

result = df.stack()
print(result)

• Line [1] creates a DataFrame with index labels and columns specified. This output saves to df.
• Line [2] outputs the DataFrame to the terminal.
• Line [3] stacks the DataFrame and saves the output to result.
• Line [4] outputs the result to the terminal (stacked format).

Output

df

	Age	Grade
Micah	8	7
Philip	7	5

result

Micah	Age	8
	Grade	7
Philip	Age	7
	Grade	5
dtype: int64

---

DataFrame unstack()

The unstack() method returns a re-shaped Multi-Level index DataFrame/Series with a new column level using inner-level labels.

#6 Pandas: Stack Unstack | Pandas Tutorial | Pandas dataframes: Reshaping, Stacking Unstacking

The syntax for this method is as follows:

DataFrame.unstack(level=- 1, fill_value=None)

Parameters	Description
level	This parameter is the level(s) to unstack. Levels can be a string, integer, or list. -1 by default (last level).
dropna	This parameter determines if rows containing missing values drop. True, by default.

We have two (2) students with relevant details that save to a DataFrame. The code below displays the original DataFrame and the DataFrame using the unstack() method.

df = pd.DataFrame([[8, 7], [7, 5]],
                  index=['Micah', 'Philip'],
                  columns=['Age', 'Grade'])
df = df.stack()
print(df)

result = df.unstack(level=0)
print(result)

• Line [1] creates a DataFrame with index labels and columns specified. The output saves to df.
• Line [2] stacks the DataFrame. The df updates accordingly.
• Line [3] outputs the DataFrame to the terminal.
• Line [4] unstacks the stacked DataFrame. The output saves to result.
• Line [5] outputs the result to the terminal (unstacked format).

Output

df

Micah	Age	8
	Grade	7
Philip	Age	7
	Grade	5
dtype: int64

result

	Age	Grade
Micah	8	7
Philip	7	5

DataFrame swap_axes()

The swapaxes() method swap axis 1 with axis 2. This parameter returns the called DataFrame/Series.

The syntax for this method is as follows:

DataFrame.swapaxes(axis1, axis2, copy=True)

axis1, axis2	If zero (0) or index is selected, apply to each column. Default is 0 (column). If zero (1) or columns, apply to each row.
copy	If True, a copy of the original DataFrame/Series creates. True, by default. If False, the updates occur on the original DataFrame/Series.

We have six (6) book titles and associated details that save to a DataFrame. The np.random.randint() method retrieves the number of books sold.

index = [('Mystery', 'Sharp Objects', 1982), ('Mystery', 'A Murder', 1973), ('Mystery', 'Wanted', 1990),
         ('Fiction', 'Thirst', 1992), ('Fiction', 'The Time Keeper', 2014), ('Fiction', 'Eligible', 1997)]
m_index = pd.MultiIndex.from_tuples(index)
num_sold = np.random.randint(5,250,size=6)

df = pd.DataFrame({'Sold': num_sold}, index=m_index)
print(df)

result = df.swapaxes(1, 0)
print(result)

• Line [1] creates a List of tuples. Each tuple contains three (3) values. The output saves to the index.
• Line [2] creates a MultiIndex from the list of tuples created on line [1] and saves it to m_index.
• Line [3] generates five (5) random integers between the specified range and saves them to num_sold.
• Line [4] creates a DataFrame from the variables created on lines [1-3] and saves to df.
• Line [5] outputs the DataFrame to the terminal.
• Line [6] swaps out the axes as specified. The output saves to result.
• Line [7] outputs the result to the terminal.

Output

df

			Sold
Mystery	Sharp Objects	1982	76
	A Murder	1973	114
	Wanted	1990	244
Fiction	Thirst	1992	153
	The Time Keeper	2014	207
	Eligible	1997	175

result

	Mystery	Fiction
	Sharp Objects	A Murder	Wanted	Thirst	The Time Keeper	Eligible
	1982	1973	1990	1992	2014	1997
Sold	76	114	244	153	207	175

---

Further Learning Resources

This is Part 14 of the DataFrame method series.

• Part 1 focuses on the DataFrame methods abs(), all(), any(), clip(), corr(), and corrwith().
• Part 2 focuses on the DataFrame methods count(), cov(), cummax(), cummin(), cumprod(), cumsum().
• Part 3 focuses on the DataFrame methods describe(), diff(), eval(), kurtosis().
• Part 4 focuses on the DataFrame methods mad(), min(), max(), mean(), median(), and mode().
• Part 5 focuses on the DataFrame methods pct_change(), quantile(), rank(), round(), prod(), and product().
• Part 6 focuses on the DataFrame methods add_prefix(), add_suffix(), and align().
• Part 7 focuses on the DataFrame methods at_time(), between_time(), drop(), drop_duplicates() and duplicated().
• Part 8 focuses on the DataFrame methods equals(), filter(), first(), last(), head(), and tail()
• Part 9 focuses on the DataFrame methods equals(), filter(), first(), last(), head(), and tail()
• Part 10 focuses on the DataFrame methods reset_index(), sample(), set_axis(), set_index(), take(), and truncate()
• Part 11 focuses on the DataFrame methods backfill(), bfill(), fillna(), dropna(), and interpolate()
• Part 12 focuses on the DataFrame methods isna(), isnull(), notna(), notnull(), pad() and replace()
• Part 13 focuses on the DataFrame methods drop_level(), pivot(), pivot_table(), reorder_levels(), sort_values() and sort_index()
• Part 14 focuses on the DataFrame methods nlargest(), nsmallest(), swap_level(), stack(), unstack() and swap_axes()
• Part 15 focuses on the DataFrame methods melt(), explode(), squeeze(), to_xarray(), t() and transpose()
• Part 16 focuses on the DataFrame methods append(), assign(), compare(), join(), merge() and update()
• Part 17 focuses on the DataFrame methods asfreq(), asof(), shift(), slice_shift(), tshift(), first_valid_index(), and last_valid_index()
• Part 18 focuses on the DataFrame methods resample(), to_period(), to_timestamp(), tz_localize(), and tz_convert()
• Part 19 focuses on the visualization aspect of DataFrames and Series via plotting, such as plot(), and plot.area().
• Part 20 focuses on continuing the visualization aspect of DataFrames and Series via plotting such as hexbin, hist, pie, and scatter plots.
• Part 21 focuses on the serialization and conversion methods from_dict(), to_dict(), from_records(), to_records(), to_json(), and to_pickles().
• Part 22 focuses on the serialization and conversion methods to_clipboard(), to_html(), to_sql(), to_csv(), and to_excel().
• Part 23 focuses on the serialization and conversion methods to_markdown(), to_stata(), to_hdf(), to_latex(), to_xml().
• Part 24 focuses on the serialization and conversion methods to_parquet(), to_feather(), to_string(), Styler.
• Part 25 focuses on the serialization and conversion methods to_bgq() and to_coo().

Also, have a look at the Pandas DataFrame methods cheat sheet!