Open in app

Sign up

Sign in

Write

Sign up

Sign in

Home

Following

Library

Stories

Stats

Mastodon

Mastering Pandas DataFrames: A Comprehensive Guide — Codes With Pankaj

12 min readSep 28, 2023

--

1. How to Create a Pandas DataFrame
   Fundamental DataFrame Operations
2. How to Select an Index or Column from a Pandas DataFrame
3. How to Add an Index, Row or Column to a Pandas DataFrame
   Adding an Indext to a DataFrame
   Adding Rows to a DataFrame
   Adding a Column to your DataFrame
   Resetting the Index of your DataFrame
4. How to Delete Indices, Rows or Columns from a Pandas Data Frame
5. How to Rename the Index or Columns of a Pandas DataFrame
6. How To Format The Data in Your Pandas DataFrame
7. How to Create an Empty DataFrame
8. Does Pandas Recognize Dates when Importing Data?
9. When, why, and how you Should Reshape your Pandas DataFrame
10. How to Iterate over a Pandas DataFrame
11. How to Write a Pandas DataFrame to a File

Pandas is a powerful data manipulation and analysis library in Python, and at the heart of it lies the DataFrame. Pandas DataFrames are versatile data structures that allow you to store, manipulate, and analyze data efficiently. In this comprehensive guide, we will explore various aspects of working with Pandas DataFrames, from creating and manipulating them to advanced operations and best practices.

How to Create a Pandas DataFrame

Creating a Pandas DataFrame is one of the fundamental steps in working with data using the Pandas library in Python. DataFrames can be created from various data sources, including dictionaries, lists, NumPy arrays, and external data files. Let’s explore different ways to create a Pandas DataFrame:

1. Creating a DataFrame from a Dictionary:

You can create a DataFrame from a Python dictionary where keys become column names, and values become the data in those columns. Here’s how you can do it :

import pandas as pd
# Create a dictionary
data = {
    'Name': ['Alice', 'Bob', 'Charlie'],
    'Age': [25, 30, 35],
    'City': ['New York', 'San Francisco', 'Los Angeles']
}
# Create a DataFrame
df = pd.DataFrame(data)
# Display the DataFrame
print(df)

This will produce the following DataFrame:

Name  Age           City
0    Alice   25       New York
1      Bob   30  San Francisco
2  Charlie   35    Los Angeles

2. Creating a DataFrame from a List of Lists:

You can also create a DataFrame from a list of lists where each inner list represents a row of data. You can specify column names separately if needed:

import pandas as pd
# Create a list of lists
data = [
    ['Alice', 25, 'New York'],
    ['Bob', 30, 'San Francisco'],
    ['Charlie', 35, 'Los Angeles']
]
# Specify column names
columns = ['Name', 'Age', 'City']
# Create a DataFrame
df = pd.DataFrame(data, columns=columns)
# Display the DataFrame
print(df)

3. Creating a DataFrame from External Data:

Pandas allows you to read data from various external sources like CSV, Excel, SQL databases, and more using functions like pd.read_csv(), pd.read_excel(), and pd.read_sql(). Here's an example of reading data from a CSV file:

import pandas as pd
# Read data from a CSV file
df = pd.read_csv('data.csv')
# Display the DataFrame
print(df)

4. Creating a DataFrame from a NumPy Array:

If you have data in a NumPy array, you can easily convert it into a DataFrame:

import pandas as pd
import numpy as np
# Create a NumPy array
data = np.array([
    ['Alice', 25, 'New York'],
    ['Bob', 30, 'San Francisco'],
    ['Charlie', 35, 'Los Angeles']
])
# Create a DataFrame from the NumPy array
df = pd.DataFrame(data, columns=['Name', 'Age', 'City'])
# Display the DataFrame
print(df)

These are some common ways to create Pandas DataFrames. Once you have a DataFrame, you can perform various operations and data manipulations on it for further analysis.

How to Select an Index or Column from a Pandas DataFrame

Selecting specific columns or rows from a Pandas DataFrame is a fundamental operation when working with data. Pandas provides several methods to do this efficiently. Let’s explore how to select an index or column from a Pandas DataFrame:

1. Selecting Columns:

You can select one or more columns from a DataFrame using square bracket notation or dot notation.

  • Using Square Bracket Notation:
import pandas as pd
# Create a sample DataFrame
data = {
    'Name': ['Alice', 'Bob', 'Charlie'],
    'Age': [25, 30, 35],
    'City': ['New York', 'San Francisco', 'Los Angeles']
}
df = pd.DataFrame(data)
# Select a single column using square brackets
name_column = df['Name']
# Select multiple columns using a list of column names
selected_columns = df[['Name', 'Age']]
# Display the selected columns
print(name_column)
print(selected_columns)
  • Using Dot Notation (for single columns with valid variable names):
# Select a single column using dot notation
name_column = df.Name
# Display the selected column
print(name_column)

2. Selecting Rows and Slices:

You can select specific rows and slices of rows using .loc[] (label-based) and .iloc[] (integer-based) indexing.

  • Using .loc[] (Label-based):
# Select a single row by label
alice_data = df.loc[0]
# Select multiple rows by label
selected_rows = df.loc[1:2]
# Select specific rows and columns
subset_data = df.loc[0:1, ['Name', 'City']]
# Display the selected rows and data
print(alice_data)
print(selected_rows)
print(subset_data)
  • Using .iloc[] (Integer-based):
# Select a single row by integer index
bob_data = df.iloc[1]
# Select multiple rows by integer indices
selected_rows = df.iloc[1:3]
# Select specific rows and columns by integer indices
subset_data = df.iloc[0:2, [0, 2]]  # Rows 0 and 1, Columns 0 and 2
# Display the selected rows and data
print(bob_data)
print(selected_rows)
print(subset_data)

3. Selecting Rows Based on Conditions:

You can also select rows based on certain conditions using boolean indexing.

# Select rows where Age is greater than 30
selected_rows = df[df['Age'] > 30]
# Select rows where City is 'San Francisco'
sf_data = df[df['City'] == 'San Francisco']
# Display the selected rows
print(selected_rows)
print(sf_data)

These are the basic techniques for selecting columns, rows, and data from a Pandas DataFrame. Depending on your data analysis needs, you can use these methods to extract the relevant information for further processing and analysis.

How to Add an Index, Row or Column to a Pandas DataFrame

In Pandas, you can add an index, row, or column to a DataFrame using various methods and operations. Let’s explore how to perform these operations:

Adding an Index to a DataFrame:

Set an Existing Column as the Index:

You can set an existing column as the index of the DataFrame using the .set_index() method:

import pandas as pd

data = {
    'Name': ['Alice', 'Bob', 'Charlie'],
    'Age': [25, 30, 35],
}

df = pd.DataFrame(data)

# Set the 'Name' column as the index
df.set_index('Name', inplace=True)

# Display the DataFrame with the new index
print(df)

Create a New Numeric Index:

If you want to create a new numeric index, you can do so using the .reset_index() method:

# Reset the index and create a new numeric index
df.reset_index(inplace=True)

# Display the DataFrame with the numeric index
print(df)

Adding Rows to a DataFrame:

  1. Using .loc[] or .iloc[]: You can add rows to a DataFrame by specifying the index label or integer index and assigning data using .loc[] or .iloc[]:
# Add a new row with label 'D' using .loc[]
df.loc['D'] = [ 'David', 28 ]

# Add a new row at a specific position (e.g., position 1)
df.loc[1] = ['Eve', 22]

# Display the DataFrame with the new rows
print(df)

Adding a Column to Your DataFrame:

  1. Using Square Bracket Notation: You can add a new column to a DataFrame using square bracket notation:
# Add a new column 'City' with data
df['City'] = ['New York', 'San Francisco', 'Los Angeles', 'Chicago']

# Display the DataFrame with the new column
print(df)

Using .insert(): You can also use the .insert() method to insert a column at a specific location (position) in the DataFrame:

# Insert a new column 'Gender' at position 2
df.insert(2, 'Gender', ['Female', 'Male', 'Male', 'Female'])

# Display the DataFrame with the new column
print(df)

Resetting the Index of Your DataFrame:

If you want to reset the index of a DataFrame to the default numeric index, you can use the .reset_index() method without specifying any arguments:

# Reset the index to the default numeric index
df.reset_index(drop=True, inplace=True)

# Display the DataFrame with the numeric index
print(df)

How to Delete Indices, Rows or Columns from a Pandas Data Frame

Deleting indices, rows, or columns from a Pandas DataFrame is a common operation when cleaning or preprocessing data. Pandas provides several methods to achieve this. Let’s explore how to delete indices, rows, and columns from a DataFrame:

Deleting Indices:

Resetting the Index : If you want to remove the current index and revert to the default numeric index, you can use the .reset_index() method:

import pandas as pd

data = {
    'Name': ['Alice', 'Bob', 'Charlie'],
    'Age': [25, 30, 35],
}

df = pd.DataFrame(data)

# Reset the index to the default numeric index
df.reset_index(drop=True, inplace=True)

# Display the DataFrame with the numeric index
print(df)

Deleting Rows :

Using Index Label: You can delete a row by specifying its index label using the .drop() method:

# Delete the row with index label '1'
df.drop(1, inplace=True)

# Display the DataFrame after deleting the row
print(df)

Using Conditional Selection: You can delete rows based on specific conditions using boolean indexing. For example, to delete rows where the ‘Age’ column is less than 30 :

# Delete rows where Age is less than 30
df = df[df['Age'] >= 30]

# Display the DataFrame after deleting rows
print(df)

Deleting Columns:

Using del Keyword: You can delete a column using the del keyword, but this operation is performed in-place:

# Delete the 'Age' column
del df['Age']

# Display the DataFrame after deleting the column
print(df)

Using .drop() Method (for non-inplace operations): To delete a column without modifying the original DataFrame, you can use the .drop() method with the axis parameter set to 1:

# Delete the 'Age' column without modifying the original DataFrame
df_without_age = df.drop('Age', axis=1)

# Display the DataFrame without the 'Age' column
print(df_without_age)

Keep in mind that the methods like .drop() that don't modify the original DataFrame return a new DataFrame with the specified changes, so it's important to assign the result back to a variable if you want to keep the modified DataFrame.

How to Rename the Index or Columns of a Pandas DataFrame

Renaming the index or columns of a Pandas DataFrame can be necessary to make your data more meaningful or to match specific naming conventions. Pandas provides methods for renaming both the index and columns. Let’s explore how to rename the index or columns of a Pandas DataFrame:

Renaming Columns:

You can rename columns using the .rename() method, which allows you to provide a mapping of the old column names to the new column names. Here's how you can do it:

import pandas as pd
data = {
    'OldName1': [1, 2, 3],
    'OldName2': [4, 5, 6],
}
df = pd.DataFrame(data)
# Create a dictionary to map old column names to new column names
new_column_names = {
    'OldName1': 'NewName1',
    'OldName2': 'NewName2',
}
# Rename the columns using the dictionary
df.rename(columns=new_column_names, inplace=True)
# Display the DataFrame with the renamed columns
print(df)

In this example, we’ve renamed ‘OldName1’ to ‘NewName1’ and ‘OldName2’ to ‘NewName2’.

Renaming the Index:

You can rename the index of a DataFrame using the .rename_axis() method. Here's how you can do it:

# Rename the index to 'NewIndex'
df = df.rename_axis('NewIndex')
# Display the DataFrame with the renamed index
print(df)

This renames the index to ‘NewIndex’.

Renaming Specific Columns or the Index by Index Position:

If you want to rename columns or the index by their position (integer index), you can use the .set_axis() method:

# Rename the first column (position 0) to 'FirstColumn'
df = df.set_axis(['FirstColumn', 'NewName2'], axis=1, inplace=False)
# Display the DataFrame with the renamed column
print(df)

In this example, we’ve renamed the first column (position 0) to ‘FirstColumn’.

Renaming All Columns or the Index Simultaneously:

If you want to rename all columns or the index simultaneously, you can directly assign a list of new names to the columns attribute for columns and index attribute for the index:

# Rename all columns simultaneously
df.columns = ['Column1', 'Column2']
# Display the DataFrame with the renamed columns
print(df)
# Rename the index simultaneously
df.index = ['Index1', 'Index2', 'Index3']
# Display the DataFrame with the renamed index
print(df)

These are the methods you can use to rename columns or the index of a Pandas DataFrame. Renaming is a crucial step in data preprocessing and analysis, allowing you to make your data more informative and compatible with your analysis requirements.

How To Format The Data in Your Pandas DataFrame

Formatting Numerical Data:

Formatting Numeric Values with Specific Decimal Places: You can format numeric columns to display a specific number of decimal places using the .round() method:

import pandas as pd

data = {
    'Value': [123.456, 456.789, 789.123],
}

df = pd.DataFrame(data)

# Format the 'Value' column to 2 decimal places
df['Value'] = df['Value'].round(2)

# Display the DataFrame with formatted values
print(df)

Adding Thousand Separators : You can add thousand separators to large numeric values for better readability using the .applymap() method with a custom formatting function:

def format_thousands(value):
    return f'{value:,}'

df = df.applymap(format_thousands)

# Display the DataFrame with formatted values
print(df)

Formatting Date Data:

  1. Parsing and Formatting Date Columns: When dealing with date columns, you can parse and format them using the pd.to_datetime() method and then apply custom formatting using the .dt.strftime() method:
import pandas as pd

data = {
    'Date': ['2023-09-01', '2023-09-15', '2023-09-30'],
}

df = pd.DataFrame(data)

# Parse the 'Date' column to datetime format
df['Date'] = pd.to_datetime(df['Date'])

# Format the 'Date' column as 'dd/mm/yyyy'
df['Date'] = df['Date'].dt.strftime('%d/%m/%Y')

# Display the DataFrame with formatted date values
print(df)

Handling Missing Data:

Replacing Missing Values: You can replace missing values in your DataFrame with specific values or by applying a fill method using the .fillna() method:

import pandas as pd
import numpy as np

data = {
    'Value': [123.456, np.nan, 789.123],
}

df = pd.DataFrame(data)

# Replace missing values with a default value (e.g., 0)
df['Value'].fillna(0, inplace=True)

# Display the DataFrame with replaced missing values
print(df)

Dropping Rows with Missing Values: Alternatively, you can drop rows containing missing values using the .dropna() method:

# Drop rows with missing values
df.dropna(inplace=True)

# Display the DataFrame after dropping missing values
print(df)

These are some common techniques to format data in a Pandas DataFrame. Depending on your specific data and analysis requirements, you can apply various formatting and cleaning operations to make your data more suitable for analysis and visualization.

How to Create an Empty DataFrame:

Creating an empty DataFrame is a common operation when you plan to populate it with data later. You can create an empty DataFrame using the Pandas library by specifying the column names or without any columns initially. Here’s how:

import pandas as pd
# Create an empty DataFrame without columns
empty_df = pd.DataFrame()
# Display the empty DataFrame
print(empty_df)

If you want to create an empty DataFrame with specified column names, you can do it like this:

# Create an empty DataFrame with column names
column_names = ['Name', 'Age', 'City']
empty_df_with_columns = pd.DataFrame(columns=column_names)

# Display the empty DataFrame with columns
print(empty_df_with_columns)

Does Pandas Recognize Dates when Importing Data?

Yes, Pandas is capable of recognizing and parsing dates when importing data from various sources, such as CSV files, Excel files, and databases. This is especially useful when your dataset contains date columns, and you want to work with them as datetime objects.

For example, when reading a CSV file using pd.read_csv(), you can specify the parse_dates parameter to indicate which columns should be treated as dates:

import pandas as pd

# Read a CSV file with a 'date_column' to be parsed as dates
df = pd.read_csv('data.csv', parse_dates=['date_column'])

# Now, 'date_column' in the DataFrame will be recognized as datetime objects

Pandas also provides functions like pd.to_datetime() to explicitly convert date strings to datetime objects when needed.

When, Why, and How You Should Reshape Your Pandas DataFrame:

When:

  • You should consider reshaping your DataFrame when it’s not in the desired format for your analysis or visualization tasks.
  • Reshaping is often necessary when working with datasets that have a wide format (many columns) and need to be transformed into a long format or vice versa.

Why:

  • Reshaping helps in making your data suitable for specific data analysis or visualization techniques.
  • It can simplify data aggregation, grouping, and reporting tasks.
  • Some machine learning models require data to be in a certain format.

How:

  • Use methods like .melt(), .pivot(), .stack(), and .unstack() to reshape your DataFrame based on your requirements.
  • Understand the concepts of “wide” and “long” format and choose the appropriate method to convert between them.

How to Iterate over a Pandas DataFrame:

You can iterate over a Pandas DataFrame using various methods, including iterrows(), itertuples(), and simple for loops. Here's how to use iterrows() as an example:

import pandas as pd

data = {
    'Name': ['Alice', 'Bob', 'Charlie'],
    'Age': [25, 30, 35],
}

df = pd.DataFrame(data)

# Using iterrows to iterate over rows
for index, row in df.iterrows():
    print(f'Index: {index}, Name: {row["Name"]}, Age: {row["Age"]}')

However, it’s important to note that iterrows() can be relatively slow for large DataFrames. In many cases, vectorized operations and functions are more efficient for data manipulation.

How to Write a Pandas DataFrame to a File:

You can write a Pandas DataFrame to various file formats using Pandas’ I/O functions. Some common output formats include CSV, Excel, SQL databases, and more. Here’s how to write a DataFrame to a CSV file as an example:

import pandas as pd

data = {
    'Name': ['Alice', 'Bob', 'Charlie'],
    'Age': [25, 30, 35],
}

df = pd.DataFrame(data)

# Write the DataFrame to a CSV file
df.to_csv('output.csv', index=False)  # Set index=False to exclude the index in the output

# You can also write to other formats, e.g., Excel or SQL, using appropriate functions

You can customize the output file’s format, delimiter, and other options according to your needs when using different output functions like .to_csv(), .to_excel(), or .to_sql().

Free

Distraction-free reading. No ads.

Organize your knowledge with lists and highlights.

Tell your story. Find your audience.

Sign up for free

Membership

Read member-only stories

Support writers you read most

Earn money for your writing

Listen to audio narrations

Read offline with the Medium app

Try for $5/month
Pandas
Pandas Dataframe
Python
Pandas Tutorial
Pandas Series

--

--

Codes With Pankaj
Codes With Pankaj
Follow

Written by Codes With Pankaj

99 Followers
4 Following

@codeawithpankaj.com

Follow

No responses yet

Write a response

What are your thoughts?

Help

Status

About

Careers

Press

Blog

Privacy

Rules

Terms

Text to speech