Application Program Interfaces (APIs)

What is an API

• Allows two pieces of software to communicate
  • E.g., your program, data, and other software components
• You interact with the API via inputs and outputs
• Abstracts the internal workings; focus on inputs/outputs

API Libraries

• Pandas API:
  • A set of software components, not all in Python
  • Used to process data by communicating with other software components
  • Example: Creating a dictionary and passing it to DataFrame constructor
    • This is an 'instance' in API terminology
    • Data in dictionary passed to Pandas API
  • Dataframe methods (head, mean) communicate with the API

REST API

• Allows communication over the internet
• Utilizes resources like storage, data access, AI algorithms
• REST: Representational State Transfer
  • Client: Your program
  • Resource: The web service
  • Endpoint: How the client finds the service
  • Involves rules for communication, input (request) & output (response)

HTTP Methods

• Transmit data over the internet
• Requests communicated via HTTP messages (often JSON files)
• Service performs operations and returns responses via HTTP messages
• Information returned usually in JSON format

Example: NBA API by Swar Patel

• Use Case: Constantly updating sports data
• NBA API always updated from endpoints at NBA.com
• Making a Request:
  • Requesting specific team info using unique id
  • Import module teams and use get_teams method
    • Returns list of dictionaries with team info
  • Convert dictionary to a table (DataFrame)
    • Use team nickname to find unique id
    • Use the id to request specific team information
• League Game Finder:
  • Makes an API call with team_id_nullable (unique team ID)
  • API call transmitted to NBA.com, response returned
  • gamefinder object’s get_data_frame method returns DataFrame
  • DataFrame includes game details
    • PLUS MINUS column indicates game outcome (positive = win, negative = loss)
    • Matchup column details opponent and game location
  • Can create DataFrames for specific matchups (e.g., home vs away games)
  • Example visualization: Plotting PLUS MINUS for home vs away games (Warriors performed better at home)