learn-co-students / api-lesson-seattle-ds-080519

APIs Lesson

APIs Lesson

Students Will Be Able To:

• Describe the client-server model and request-response cycle
• Identify examples of clients and servers, both on their local computers and remotely connected
  • PostgreSQL is a server
  • psql in the command line is a client
  • psycopg2 in Python is a client
  • Jupyter Notebook is a server
  • Google Chrome (like all web broswers) is a client
• Identify examples of requests and responses
  • If you're sending a request to PostgreSQL, that looks like "SELECT * FROM names;"
  • If you're sending a request to Jupyter Notebook, that looks like "http://localhost:8888/notebooks/APIs_exploratory_notebook.ipynb"
  • If you're sending a request GitHub to get their homepage, that looks like "github.com"
• Understand the basic concept of a web server, and the types of requests and responses it uses
• Understand the difference between a request that returns HTML and an API request that returns JSON
  • "API" technically can mean a lot more things, but usually it means an HTTP (web) interface that returns JSON
• Understand the basic pieces of making HTML and JSON requests
  • For simple requests, you only need 2 things: HTTP verb and URL
  • HTTP verb: this will almost always be GET, kind of like how a data scientist's SQL query will almost always be SELECT. But there are other verbs, including POST and PATCH, which are like INSERT INTO and UPDATE
  • URL: this is the string representing the location of the resource. It can also be called a "path", or an "endpoint"
• Use a browser to make HTML and JSON requests
  • This just means typing the URL into the browser. It will always be a GET request.
  • The browser will show HTML or JSON depending on the URL. So, reddit.com shows you HTML, whereas reddit.com/.json shows you JSON.
• Use Postman to make HTML and JSON requests
  • Postman has input fields where you can type the URL and select the HTTP verb from a drop-down
  • Remember that this is a good tool for exploration/figuring out how to use an API (better than Chrome), but it's not the final product
• Practice reading someone else's Python code that uses the requests library
  • We walked through the Google Books example (see "Existing Code" section below)
• Use Python requests library to make JSON requests
  • This looks like requests.get("http://api.open-notify.org/iss-now.json")
  • The HTTP verb is specified by the method name (.get for GET, .post for POST, etc.)
  • The URL is the first (and only required) parameter passed into the .get method
• Practice reading the documentation for an API

Existing Code (google_books.py)

1. Uses the Python requests library
2. Uses the Google Books API
3. Searches for author "Jake VanderPlas" (author of my go-to Python data science book)
4. Prints out a list of book titles that match this query

Deliverables

• [] Refactor the existing Google Books code so that it uses the params argument (we did not get to this in class)
• Write new code in a new file that queries the NASA ISS API and prints out the current latitude and longitude
  • This is in APIs_exploratory_notebook.ipynb, plus some fun mapping stuff. You might need to run conda install ipyleaflet in the terminal to get it working. If you run it again, expect a different map, since the ISS is constantly moving very quickly!
  • Currently the code in the notebook doesn't do any error checking. That is pretty common for exploratory code in a notebook, but you almost always want to do error checking when you write actual code that uses an API