This repository contains the code and analysis focusing on mortality data from the United States.
Link to dataset (https://www.kaggle.com/cdc/mortality) The project utilizes a dataset named "Death in the United States," consisting of:
CSV files: Each file contains data for a specific year, with various attributes related to deaths.
JSON files: Each JSON file corresponds to a CSV file and provides code mappings for interpreting the data.
code/: Contains Python scripts for data analysis and machine learning tasks.
data/: Stores the CSV and JSON files from the original dataset.
results/: Intended to hold output files, such as Excel sheets for visualizations.
Data Exploration:
Scripts in code/ perform analysis on individual variables (e.g., "manner of death") and combinations of variables (e.g., "manner of death" with "sex").
Analysis results are intended to be exported to results/ for further exploration and visualization.
Machine Learning:
Scripts in code/ aim to build a model to predict accidental deaths.
Features like education, age, place of death, and others are used for training.
The Random Forest Classifier is a potential candidate for this task.
This project requires the Apache Spark library for data processing. You can install it using pip install pyspark.
Running the Analysis:
Ensure you have Spark installed.
Download or access the dataset (CSV and JSON files).
Place the dataset files in the data/ directory.
Execute the desired Python scripts from the code/ directory.
Analysis results might be exported to results/ for further exploration.