Stat-Analyzer

PhD-level statistical analysis engine for structured and semi-structured data, focused purely on statistics, probability, and decision science.

Stat Analyzer 📊🧠

Stat Analyzer is a PhD-level statistical analysis and decision science engine designed to work with structured and semi-structured data.
It focuses purely on Statistics and Probability, delivering deep reasoning, diagnostics, and decision-oriented insights.

🚫 No machine learning model performance
✅ 100% statistics, inference, probability, and decision science

---

🎯 Purpose

Stat Analyzer is built to behave like a PhD-educated statistician that can:

• Understand messy real-world data
• Decide which statistical techniques apply
• Explain why they apply
• Interpret results correctly
• Translate statistics into business and research decisions

---

🔍 What This Project Covers

📌 Descriptive Statistics

• Mean, Median, Mode
• Variance, Standard Deviation
• Range, IQR
• Skewness, Kurtosis
• Frequency distributions

📌 Probability & Uncertainty

• Probability rules
• Conditional probability
• Bayes theorem
• Permutations & combinations
• Certainty vs uncertainty reasoning

📌 Statistical Distributions

• Normal
• Binomial
• Poisson
• Uniform
• Exponential
• Chi-square
• Student’s t
• F distribution

📌 Inferential Statistics

• Central Limit Theorem
• Confidence intervals
• Hypothesis testing
• P-values
• Type I & Type II errors

📌 Statistical Tests

• Z-test
• One-sample & two-sample t-tests
• Paired t-test
• ANOVA
• F-test
• Chi-square test
• Non-parametric alternatives

📌 Regression & Relationships

• Correlation analysis
• Simple & multiple linear regression
• Ordinary Least Squares (OLS)
• Residual diagnostics
• Omnibus test
• Assumption validation

📌 Error & Diagnostic Analysis

• Mean Error
• Mean Absolute Error
• Mean Percentage Error
• Absolute & percentage errors
• Error-based decision guidance

📌 Analytics Types

• Descriptive analysis
• Diagnostic analysis
• Predictive analysis (statistics-only)
• Prescriptive analysis

---

📂 Supported Data Formats

• CSV files
• Excel files
• Semi-structured text tables
• Mixed data types (int, float, categorical, missing values, outliers)

---

🧪 Sample Datasets

Sample datasets are available in the /data directory:

• stat_analyzer_test.csv
• stat_analyzer_financial_risk.csv
• stat_analyzer_semi_structured.txt

These datasets are intentionally designed with:

• Missing values
• Outliers
• Mixed numeric types
• Real-world inconsistencies

---

🖥️ Application Behavior

The system automatically:

• Detects data types
• Identifies missing values and outliers
• Checks statistical assumptions
• Recommends appropriate tests
• Explains results in simple and advanced terms
• Links statistical outcomes to decisions

---

🚀 How to Run (Basic)

pip install -r requirements.txt
python app/main.py