📊 SAS Data Analysis & Modeling Projects

This repository contains a collection of end-to-end SAS analytics projects showcasing skills in data preparation, visualization, statistical analysis, and predictive modeling.
Each project applies SAS programming to real-world-style datasets from SASHELP and external sources, covering topics from operational HR analytics to sports statistics, retail data, and nutritional analysis.

---

📁 Projects Overview

🔹 End-to-End Data Analytics Project

• Multi-dataset analysis demonstrating SAS statistical capabilities
• Cholesterol-by-gender visualization using vertical bar charts
• BMI computation and distribution analysis
• Correlation analysis for sports performance metrics
• Simple and multiple regression modeling with diagnostics
• Multicollinearity detection and backward elimination for model refinement

🔹 Firefighter Overtime Pay Analysis

• Created categorical OT pay variables
• Conducted binomial proportion tests
• Performed Chi-square tests for independence
• Used one-way ANOVA with Tukey’s post-hoc comparisons

🔹 Baseball Player Salary Modeling (SASHELP.Baseball)

• Built simple and multiple regression models for salary prediction
• Evaluated diagnostics (residuals, VIF)
• Applied stepwise regression for optimal predictor selection

🔹 Car Price Prediction (SASHELP.CARS)

• Developed multiple regression models for MSRP
• Identified and addressed multicollinearity
• Applied backward elimination to improve model stability

🔹 SASHELP.Shoes Sales Visualization

• Created product sales bar charts and sales distribution boxplots
• Designed 3D pie charts for inventory composition
• Enhanced graphical output with sorting, color customization, and labels

🔹 PCA Analysis – Pizza Nutritional Data

• Performed multicollinearity checks
• Applied Principal Components Analysis (PCA)
• Compared full vs reduced models for efficiency and interpretability

---

🧠 Key Skills Demonstrated

• Data Preparation: DATA step transformations, recoding variables
• Statistical Testing: ANOVA, Chi-square, Binomial tests
• Regression Modeling: simple, multiple, and stepwise regression
• Dimensionality Reduction: PCA for multicollinearity mitigation
• Visualization: PROC SGPLOT, PROC GCHART, PROC BOXPLOT
• Diagnostics & Model Selection: R², p-values, VIF, residual plots

---

⚙️ Tools & Resources

• SAS 9.4 / SAS Studio
• Built-in datasets from the SASHELP library
• Custom-created datasets for applied problem solving

---

👤 Author

Aryan Sharma
Data Analytics | SAS Programming | Predictive Modeling

---

📜 License

This repository is intended for educational and portfolio purposes.
If you use or adapt this work, please give proper credit.