AI Data Analysis Assistant: Let Data Speak

class DataAnalysisAI:
    """Intelligent Data Analysis Assistant"""
    
    def analyze_dataset(self, df, business_context=""):
        """Comprehensive dataset analysis"""
        # 1. Data overview
        overview = self.data_overview(df)
        
        # 2. Use LLM to understand data
        data_understanding = self.llm.generate(f"""
        Dataset information:
        - Shape: {df.shape}
        - Columns: {list(df.columns)}
        - Data types: {df.dtypes.to_dict()}
        - Sample data: {df.head().to_string()}
        
        Business context: {business_context}
        
        Please analyze:
        1. Which business scenarios could this dataset support?
        2. Which fields are most important?
        3. What data quality issues may exist?
        """)
        
        # 3. Statistical analysis
        stats = self.statistical_analysis(df)
        
        # 4. Discover insights
        insights = self.discover_insights(df, stats)
        
        # 5. Generate report
        report = self.generate_report(
            overview, stats, insights, data_understanding
        )
        
        return report
    
    def discover_insights(self, df, stats):
        """Discover data insights"""
        insights = []
        
        # Trend analysis
        if self.has_time_column(df):
            trends = self.analyze_trends(df)
            insights.extend(trends)
        
        # Correlation analysis
        correlations = self.find_correlations(df)
        insights.extend(correlations)
        
        # Anomaly detection
        anomalies = self.detect_anomalies(df)
        insights.extend(anomalies)
        
        # In-depth analysis with LLM
        llm_insights = self.llm.generate(f"""
        Based on the following statistics, discover business insights:
        {json.dumps(stats, indent=2)}
        
        Please provide:
        1. 3-5 key findings
        2. Potential business implications
        3. Recommended follow-up analysis directions
        """)
        
        insights.append(llm_insights)
        return insights
    
    def generate_sql_query(self, question, schema):
        """Natural language to SQL"""
        prompt = f"""
        Database schema:
        {schema}
        
        User question: {question}
        
        Please generate an SQL query.
        """
        
        sql = self.llm.generate(prompt)
        return sql
    
    def explain_analysis(self, results, for_audience="business"):
        """Explain analysis results"""
        if for_audience == "business":
            prompt = f"""
            Explain the following technical analysis results in business language:
            {results}
            
            Requirements:
            1. Avoid technical jargon
            2. Highlight business value
            3. Provide actionable recommendations
            """
        else:
            prompt = f"""
            Explain the analysis methods and results in detail:
            {results}
            
            Including:
            1. Statistical methods used
            2. Confidence in the results
            3. Limitations
            """
        
        return self.llm.generate(prompt)