An easy-to-follow and comprehensive guide to creating data apps with Streamlit, including how-to guides for working with cloud data warehouses like Snowflake, using pretrained Hugging Face and OpenAI models, and creating apps for job interviews.
Key Features
Create machine learning apps with random forest, Hugging Face, and GPT-3.5 turbo models
Gain an insight into how experts harness Streamlit with in-depth interviews with Streamlit power users
Discover the full range of Streamlit’s capabilities via hands-on exercises to effortlessly create and deploy well-designed apps
Book Description
If you work with data in Python and are looking to create data apps that showcase ML models and make beautiful interactive visualizations, then this is the ideal book for you. Streamlit for Data Science, Second Edition, shows you how to create and deploy data apps quickly, all within Python. This helps you create prototypes in hours instead of days!
Written by a prolific Streamlit user and senior data scientist at Snowflake, this fully updated second edition builds on the practical nature of the previous edition with exciting updates, including connecting Streamlit to data warehouses like Snowflake, integrating Hugging Face and OpenAI models into your apps, and connecting and building apps on top of Streamlit databases. Plus, there is a totally updated code repository on GitHub to help you practice your newfound skills.
You'll start your journey with the fundamentals of Streamlit and gradually build on this foundation by working with machine learning models and producing high-quality interactive apps. The practical examples of both personal data projects and work-related data-focused web applications will help you get to grips with more challenging topics such as Streamlit Components, beautifying your apps, and quick deployment.
By the end of this book, you'll be able to create dynamic web apps in Streamlit quickly and effortlessly.
What you will learn
Set up your first development environment and create a basic Streamlit app from scratch
Create dynamic visualizations using built-in and imported Python libraries
Discover strategies for creating and deploying machine learning models in Streamlit
Deploy Streamlit apps with Streamlit Community Cloud, Hugging Face Spaces, and Heroku
Integrate Streamlit with Hugging Face, OpenAI, and Snowflake
Beautify Streamlit apps using themes and components
Implement best practices for prototyping your data science work with Streamlit
Who this book is for
This book is for data scientists and machine learning enthusiasts who want to get started with creating data apps in Streamlit. It is terrific for junior data scientists looking to gain some valuable new skills in a specific and actionable fashion and is also a great resource for senior data scientists looking for a comprehensive overview of the library and how people use it. Prior knowledge of Python programming is a must, and you’ll get the most out of this book if you’ve used Python libraries like Pandas and NumPy in the past.
Author(s): Tyler Richards
Edition: 2
Publisher: Packt
Year: 2023
Language: English
Pages: 359
Preface
Who this book is for
What this book covers
Acknowledgment
To get the most out of this book
Get in touch
An Introduction to Streamlit
Technical requirements
Why Streamlit?
Installing Streamlit
Organizing Streamlit apps
Streamlit plotting demo
Making an app from scratch
Using user input in Streamlit apps
Finishing touches – adding text to Streamlit
Summary
Uploading, Downloading, and Manipulating Data
Technical requirements
The setup – Palmer’s Penguins
Exploring Palmer’s Penguins
Flow control in Streamlit
Debugging Streamlit apps
Developing in Streamlit
Exploring in Jupyter and then copying to Streamlit
Data manipulation in Streamlit
An introduction to caching
Persistence with Session State
Summary
Data Visualization
Technical requirements
San Francisco Trees – a new dataset
Streamlit visualization use cases
Streamlit’s built-in graphing functions
Streamlit’s built-in visualization options
Plotly
Matplotlib and Seaborn
Bokeh
Altair
PyDeck
Configuration options
Summary
Machine Learning and AI with Streamlit
Technical requirements
The standard ML workflow
Predicting penguin species
Utilizing a pre-trained ML model in Streamlit
Training models inside Streamlit apps
Understanding ML results
Integrating external ML libraries – a Hugging Face example
Integrating external AI libraries – an OpenAI example
Authenticating with OpenAI
OpenAI API cost
Streamlit and OpenAI
Summary
Deploying Streamlit with Streamlit Community Cloud
Technical requirements
Getting started with Streamlit Community Cloud
A quick primer on GitHub
Deploying with Streamlit Community Cloud
Debugging Streamlit Community Cloud
Streamlit Secrets
Summary
Beautifying Streamlit Apps
Technical requirements
Setting up the SF Trees dataset
Working with columns in Streamlit
Exploring page configuration
Using Streamlit tabs
Using the Streamlit sidebar
Picking colors with a color picker
Multi-page apps
Editable DataFrames
Summary
Exploring Streamlit Components
Technical requirements
Adding editable DataFrames with streamlit-aggrid
Creating drill-down graphs with streamlit-plotly-events
Using Streamlit Components – streamlit-lottie
Using Streamlit Components – streamlit-pandas-profiling
Interactive maps with st-folium
Helpful mini-functions with streamlit-extras
Finding more Components
Summary
Deploying Streamlit Apps with Hugging Face and Heroku
Technical requirements
Choosing between Streamlit Community Cloud, Hugging Face, and Heroku
Deploying Streamlit with Hugging Face
Deploying Streamlit with Heroku
Setting up and logging in to Heroku
Cloning and configuring our local repository
Deploying to Heroku
Summary
Connecting to Databases
Technical requirements
Connecting to Snowflake with Streamlit
Connecting to BigQuery with Streamlit
Adding user input to queries
Organizing queries
Summary
Improving Job Applications with Streamlit
Technical requirements
Using Streamlit for proof-of-skill data projects
Machine learning – the Penguins app
Visualization – the Pretty Trees app
Improving job applications in Streamlit
Questions
Answering Question 1
Answering Question 2
Summary
The Data Project – Prototyping Projects in Streamlit
Technical requirements
Data science ideation
Collecting and cleaning data
Making an MVP
How many books do I read each year?
How long does it take for me to finish a book that I have started?
How long are the books that I have read?
How old are the books that I have read?
How do I rate books compared to other Goodreads users?
Iterative improvement
Beautification via animation
Organization using columns and width
Narrative building through text and additional statistics
Hosting and promotion
Summary
Streamlit Power Users
Fanilo Andrianasolo
Adrien Treuille
Gerard Bentley
Arnaud Miribel and Zachary Blackwood
Yuichiro Tachibana
Summary
Other Books You May Enjoy
Index
