Readme

What

This codebase is a Heroku hosted Flask app that pulls census, given a correct POST passing in a census American Community Survey (ACS) 5 year study year (2012, 2015, 2018, etc), and updates a Google Sheets worksheet. On update of this worksheet,a seperate Tableau dashboard is updated showing vizualitions of the current Sheets values.

Who

Taylor Malory: taymal1987@gmail.com
Matthew Barnes: mlbarnes04@gmail.com

Getting Started

• Make sure you have python 3 (Heroku is currently running 3.8), Anaconda (>= 4.6) and Git.

• Clone and cd into the repo

• Create a new virtual enviroment with conda create --name myenv && conda activate myenv, substituting myenv with whatever enviroment name you choose.

• Install dependencies with conda install --file requirements.txt

• For local development - You will need a client_secret.json. This is found through here, under the account associated with this project. Click on the project (located in the upper left near the Google APIs logo), click Crednetials, click OAuth 2.0 Client IDs, download the JSON file and change the file name to just client_secret.json. Place JSON file in the root of the directory.

• For local development - In the directory you will find a .env.example file. Run cp .env.example .env, then open .env in your prefered editor. The keys present in this file are correct but their values are not. Go to Heroku, select either the staging or production project, go find the enviroment variables and replace the values in the .env file with the associated values found on Heroku.

• For local development - In your console, run FLASK_ENV=dev.

• You are now ready to edit the files and push to Heroku/development, so long as you have access to the project.

Example request

• An example POST looks like https://coic-housing.herokuapp.com/update_gsheet?pword=XXXXX&year=2018. Substituting XXXXX for the correct password. The year should be within 2013 - the year of the most current ACS.

Tips

• Ensure you have proper Heroku enviroment variables configured. These are denoted in the code with os.environ['SOME_ENV_VAR'], except for the pygsheets authorization which just needs the key name.
• The enviroment variables should be good to go and not need changing or updating. If they do, for some reason, those can be updated in Heroku either through the CLI or site. The 'SERVICE_ACCOUNT' enviroment variable is a JSON file, generated through the projects Google Developers page (https://console.developers.google.com/), and thrown directly into Heroku as an eviroment variable. If, for some reason, the project needs a new one, generate it through the above link, add the entire file as an enviroment variable, and then share access to the email address within the JSON file from the projects sheets page.
• To view available conda virtual enviroments, run conda env list
• To leave the current conda virtual enviroment, run conda deactivate

App Structure

This app uses 4 different enviroments: development, testing, staging, and production. The home route ('/') is used for capturing and sending parameters to '/update_gsheets'. The general flow of the app is as follows:

1. Recieve a GET resquest at /update_gsheet.
2. Validate user provided password and year.
3. Authorize with Google API using service account credentials. What get's passed for authorization depends on which enviroment you are in. Development and testing use a .env, while staging and production use environment variables stored in Heroku.
4. Open Google Sheets workbook. There are 4 seperate workbooks, each corresponding to a seperate environment. The service credentials, and the sharing of the email address (found within the service account file) from Google Sheets, dictates which workbook is opened.
5. Build Census query string and perform request.
6. Put response in Pandas dataframe and transform data.
7. Open worksheet.
8. Clear worksheet.
9. Write to worksheet.
10. Repeat for every worksheet.
11. Render dashboard (update can take up to 24 hours to become visible unless forced from the Tableau Public website (not app)).

Links:

• Anaconda: https://docs.anaconda.com/anaconda/install/
• Python: https://www.python.org/downloads/
• Pandas: https://pandas.pydata.org/
• pygsheets: https://pypi.org/project/gsheets/
• requests: https://requests.readthedocs.io/en/master/