rrgayhart/queryable_api

Queryable API Challenge

An internal API for a national healthcare provider (a code challenge).

Deployed Application

Accessing the API

Parameter	Description
max_discharges	The maximum number of Total Discharges
min_discharges	The minimum number of Total Discharges
max_average_covered_charges	The maximum Average Covered Charges
min_average_covered_charges	The minimum Average Covered Charges
max_average_medicare_payments	The maximum Average Medicare Payment
min_average_medicare_payments	The minimum Average Medicare Payment
state	The exact state that the provider is from

GET /providers?max_discharges=70&min_discharges=68&max_average_covered_charges=50000&min_average_covered_charges=45000&min_average_medicare_payments=6000&max_average_medicare_payments=10000&state=GA

{
  "meta": {
    "total": 2,
    "max_shown": 100,
    "page": "1"
  },
  "data": [
    {
      "Provider Name": "SPALDING REGIONAL MEDICAL CENTER",
      "Provider Street Address": "601 SOUTH 8TH STREET",
      "Provider City": "GRIFFIN",
      "Provider State": "GA",
      "Provider Zip Code": "30223",
      "Hospital Referral Region Description": "GA - Atlanta",
      "Total Discharges": 68,
      "Average Covered Charges": "$45,824.38",
      "Average Total Payments": "$9,260.82",
      "Average Medicare Payments": "$8,049.48"
    },
    {
      "Provider Name": "EMORY EASTSIDE MEDICAL CENTER",
      "Provider Street Address": "1700 MEDICAL WAY",
      "Provider City": "SNELLVILLE",
      "Provider State": "GA",
      "Provider Zip Code": "30078",
      "Hospital Referral Region Description": "GA - Atlanta",
      "Total Discharges": 68,
      "Average Covered Charges": "$46,777.98",
      "Average Total Payments": "$8,796.27",
      "Average Medicare Payments": "$7,899.88"
    }
  ]
}

Getting Started

Prerequisites

You will need to have Rails 5 and Ruby 2.4.1 and PostgreSQL up and running in order to run this application locally.

If you're unfamiliar with Ruby or Rails, check out the Rails Getting Started Documentation

Installing

Pull down this repository

git@github.com:rrgayhart/queryable_api.git

Install dependencies

bundle install

Set up the database

Note: running rake db:seed will require an internet connection and will take some time.

rake db:create
rake db:migrate
rake db:test:prepare

Start the server

rails s

Testing

Run the tests

rspec

Implementation Notes

Technology Used

• Language
  • Ruby 2.4.1
• Framework
  • Rails ~> 5.1.2
• Database
  • PostgreSQL through the pg gem (still using Active Record as the ORM)
• Testing
  • rspec through the rspec-rails gem
  • factory_girl_rails
  • shoulda-matchers
• API
  • jbuilder
  • will-paginate

Provider Schema

Data was provided in the form for this Data Set CSV

    create_table :providers do |t|
      t.string :provider_name
      t.string :provider_street_address
      t.string :provider_city
      t.string :provider_state
      t.string :provider_zip_code
      t.string :hospital_referral_region_description
      t.integer :total_discharges
      t.decimal :average_covered_charges, :precision => 10, :scale => 2, :default => 0.0
      t.decimal :average_total_payments, :precision => 10, :scale => 2, :default => 0.0
      t.decimal :average_medicare_payments, :precision => 10, :scale => 2, :default => 0.0
      t.timestamps
    end

As there was a direct match between the naming of things like :provider_street_address as a header in the CSV and Provider Street Address as a key in the expected output - I chose to keep that naming convention in the database columns.

If this project were to expand, I would consider:

• refactoring out references to 'provider' in table names
• spliting address into it's own table
• storing provider_id for deduplication accross multiple CSVs

Deploying to Heroku

I chose to try and seed the CSV data directly from the link provided rather than including the CSV in the project git history.

I was able to run the CSV seeding process in a detached state to try and avoid timeouts using: heroku run:detached rake db:seed.

Deviations from the Requirements

Pagination

Given the large quantity of records, a general query could return 163065 records. This would be an enormous response, so I paginated the responses and allowed the query to include a page parameter. Top level keys on the API responses are 'data' and 'meta' to reflect the total number of records/current page/etc.

Endpoints

The spec requires /providers endpoint (which is present) - but I namespaced the controller and view under api/v1 and provided a namespaced endpoint. I also routed the root to the providers index so as not to confuse anyone visiting the heroku site. A more elegant solution would probably have been to provide a splash page for the API with instructions for use and links to provided endpoints at the root.