An Analysis of Diabetes Health Indicators Using SAS

Date published: 10 June, 2024

SAS | Data Analysis | Statistical Analysis | Statistical Modeling

Overview

This project analyses a comprehensive diabetes health indicator dataset from a CDC survey using SAS. The dataset includes various health-related variables such as blood pressure, cholesterol levels, smoking status, and more. The results from the analysis can be found in my GitHub Repo.

Script Description

Data Read-in

Reads the diabetes health indicator dataset from a CSV file and performs initial data checks.

Initial Data Checks

Prints the schema of the data
Displays sample observations
Prints row count in the log file

Missing Data, Outliers, and Inconsistencies

Checks for missing values
Computes univariate statistics

Data Imbalance

Computes frequencies of the target variable
Displays bar plot of the target variable

Correlation Analysis

Computes correlation between variables
Macro to find highly correlated variables

Exploring Variable Relationships

Analyzes relationships between variables using chi-square tests

Data Analysis & Transformations

Merges minority groups in the target variable
Analyzes and bins age groups
Checks the distribution of BMI
Analyzes variance of BMI across age groups
Performs cluster analysis

Diabetes Prediction

Performs logistic regression to predict diabetes
Analyzes residuals
Includes interaction terms in the model

Acknowledgments

SAS documentation for providing comprehensive guides on SAS programming.
The open-source community for fostering collaboration and knowledge sharing.

References

Building Risk Prediction Models for Type 2 Diabetes Using Machine Learning Techniques

Behavioral Risk Factor Surveillance System - Data

Relevant Links:

Project Thumbnail Image Source

Page updated

Google Sites

Report abuse