Skip to content

Latest commit

 

History

History
124 lines (101 loc) · 5.67 KB

README.md

File metadata and controls

124 lines (101 loc) · 5.67 KB
Raw

Healthcare Workforce API Server

An API server for the Healthcare Workforce Needs application.

Table of Contents

  1. Development Environment Quick Start
  2. Deploying In Production
  3. Saving And Loading Model Data Into/From Fixtures
  4. Route Documentation

Development Environment Quick Start

This project uses pipenv for python package version management, so make sure you have that installed. If you need instructions for setting it up, check here. Once pipenv is installed, you can set up a virtual environment and install all python dependencies with pipenv install.

There is one final dependency that you'll need to run the application, GLPK. On mac, you can install this package with brew install glpk, but on other OSes it might be a little more tricky. Start here if you're looking for installation instructions.

Now, copy the .env.default to .env using cp .env.default .env.

Now run MySQL through docker using:

docker stop workforce-mysql
docker rm workforce-mysql

# Replace these vars when doing production work look at .env.prod
docker run \
  --name workforce-mysql \
  -e MYSQL_ROOT_PASSWORD=password \
  -e MYSQL_USER=workforceuser \
  -e MYSQL_PASSWORD=password \
  -e MYSQL_DATABASE=workforcewebapp \
  -v /home/ubuntu/workforce-db-data/:/var/lib/mysql \
  -d \
  -p 3306:3306 \
  mysql:5

Now you'll need to initialize the database and load the data in. To do so, run pipenv run migrate, and then pipenv run loadmodels.

Once pipenv is set up and the .env file is set correctly, run pipenv run serve to run a local development server at http://localhost:8000/.

Deploying In Production

See top level README.

Saving And Loading Model Data Into/From Fixtures

We provide fixture data that can be loaded into the database to initialize the application with some models. This data was generated by hand, but the data can be generated from the live DB in cases that you want to save the state of the database.

To save the live database to your local machine, run pipenv run dumpmodels.

To load that same data, replace workforceModels.json in the workforceAPI/workforceAPI/fixtures directory with your new file (and rename it to workforceModels.json). Then run pipenv run loadmodels.

Route Documentation

There are several routes set up for accessing the model data. Here are the names, allowed methods, parameters, and descriptions:

  • Name: /api

    • Allowed Methods: GET
    • Parameters: None
    • Description: Base API endpoint. Returns text and a 200 to verify everything is working. Doesn't return data.
    • Example: 
      curl 'localhost:8000/api'

  • Name:/api/file-upload

    • Allowed Methods: POST
    • Parameters:
      • metadata: a json serializable object containing the following required fields:
        model_name: Model name, used in visualization. description
        model_type: One of: "ideal_staffing", "ideal_staffing_current", or "service_allocation".
        start_year: Currently, one of: 2014, 2015, 2016, 2017, 2018, 2019, 2020, 2021, 2022, 2023, 2024.
        end_year: Currently, one of: 2014, 2015, 2016, 2017, 2018, 2019, 2020, 2021, 2022, 2023, 2024.
        step_size: The step size for the model in years.
        removed_professions: List of removed professions or [].
      • file: The data file that the model will use to compute supply and needs.
    • Description:
    • Return: Either "File uploaded successfully" and 201 if the upload succeeded, a 400 with a reason (usually missing parameters or file), or 500 for a model issue.
    • Example (must be run from the project root directory): 
      curl \
  -X POST \
  -F 'metadata={"model_name": "new_model", "description": "a model", "model_type": "ideal_staffing", "start_year": 2019, "end_year": 2020, "step_size": 1, "removed_professions": []}' \
  -F 'file=@workforceAPI/workforceAPI/test_data/Workforce_Optimization_Tool_-_Input_Data.xlsx' \
  'localhost:8000/api/file-upload'

  • Name: /api/models

    • Allowed Methods: GET
    • Parameters: None
    • Description: Loads all stored model metadata.
    • Return: JSON object, array of objects of model metadata.
    • Example: 
      curl 'localhost:8000/api/models'

  • Name:/api/rerun-model

    • Allowed Methods: POST
    • Parameters:
      • Required:
        model_id: A model id for the model that will be re-run.
        model_name: Model name, used in visualization.
        description
      • Optional:
        model_type: One of: "ideal_staffing", "ideal_staffing_current", or "service_allocation".
        start_year: Currently, one of: 2014, 2015, 2016, 2017, 2018, 2019, 2020, 2021, 2022, 2023, 2024.
        end_year: Currently, one of: 2014, 2015, 2016, 2017, 2018, 2019, 2020, 2021, 2022, 2023, 2024.
        step_size: The step size for the model in years.
        removed_professions: Comma separated list (without spaces) of professions to remove.
    • Description:
    • Return: Either "File uploaded successfully" and 201 if the model runs successfully. 400 for a bad request (missing model_id) or 500 for a failure in running the model.
    • Example: 
      curl \
  -X POST \
  -F 'model_id=fdbb2a92-4531-4395-91d2-1b495b6566d5' \
  -F 'removed_professions=NP,MA' \
  -F 'model_name=updated model NP removed to remove NP + MA' \
  -F 'description=test_update, remove NP + MA' \
  'localhost:8000/api/rerun-model'