MA Community Scores Development Project

Setup and Running Code

Install Requirements

1. Run pip freeze > requirements.txt
2. Run pip install -r requirements.txt

Run Code

1. After setting file variables in run.py, run the code using python run.py
2. For visualizations of correlation matrices, bar charts for health scores for Top 3 and Bottom 3 communities, python visualize.py
3. For visualizations of specific communities, python visualize_communities.py

data

Folder that contains all data that is used as the input for this process.

output

Folder that contains files that reflect the different outputs for this process. This files only contain numbers and values, they do not convey semantics.

visualizations

Folder that contains visualizations of the different files within the output folder.

result_projections

Folder that contains D3 code to visualize the results of the composite Health Score, and project them onto a map of the state of Massachusetts.

build_fulldataset.py

Combines determinant data and outcome data into one data set that contains all the data.

clean.py

Cleans the given dataset by converting totals to rates/percentages, removing sparse columns, normalizing percentages, and coerce strings with special characters to numbers. (i.e. "1,000,000+" => 1,000,000)

impute.py

Imputes the missing values within a given dataset.

pca.py

Establishes the HealthScores class. This class is responsible for all the functions related to producing composite health scores based off a PCA process. Factor and Correlation Analysis are also present within the class.

run.py

Produces the results seen in the output folder using files from the data folder as input.

visualize_communities.py

Produces two histograms (one for subsets, one for domains) displaying results across all different data types of data for a specific given community.

visualize.py

Produces the rest of the visualizations within the visualizations folder except for the ones made within results_projections. Correlation matricies, plotting comparisons, and top 3 vs bottom 3 comparisons.