victor-dataanalysis 0.5.6__py3-none-any.whl

victor_dataanalysis/__init__.py

@@ -0,0 +1,27 @@
+"""Data Analysis Vertical Package - Complete implementation with extensions.
+
+Competitive use case: ChatGPT Data Analysis, Claude Artifacts, Jupyter AI, Code Interpreter.
+
+This vertical provides:
+- Data exploration and profiling
+- Statistical analysis and visualization
+- Machine learning model training
+- Report generation with insights
+- CSV/Excel/JSON data processing
+"""
+
+from victor_dataanalysis.assistant import DataAnalysisAssistant
+from victor_dataanalysis.prompts import DataAnalysisPromptContributor
+from victor_dataanalysis.mode_config import DataAnalysisModeConfigProvider
+from victor_dataanalysis.safety import DataAnalysisSafetyExtension
+from victor_dataanalysis.tool_dependencies import DataAnalysisToolDependencyProvider
+from victor_dataanalysis.capabilities import DataAnalysisCapabilityProvider
+
+__all__ = [
+    "DataAnalysisAssistant",
+    "DataAnalysisPromptContributor",
+    "DataAnalysisModeConfigProvider",
+    "DataAnalysisSafetyExtension",
+    "DataAnalysisToolDependencyProvider",
+    "DataAnalysisCapabilityProvider",
+]

victor_dataanalysis/assistant.py

@@ -0,0 +1,189 @@
+"""Data Analysis Assistant - Complete vertical for data exploration and insights.
+
+Competitive positioning: ChatGPT Data Analysis, Claude Artifacts, Jupyter AI.
+"""
+
+from typing import Any, Dict, List, Optional
+
+from victor.core.verticals.base import StageDefinition, VerticalBase
+from victor.core.verticals.protocols import (
+    ModeConfigProviderProtocol,
+    PromptContributorProtocol,
+    SafetyExtensionProtocol,
+    TieredToolConfig,
+    ToolDependencyProviderProtocol,
+)
+
+# Phase 3: Import framework capabilities
+from victor.framework.capabilities import FileOperationsCapability
+
+
+class DataAnalysisAssistant(VerticalBase):
+    """Data analysis assistant for exploration, visualization, and insights.
+
+    Competitive with: ChatGPT Data Analysis, Claude Artifacts, Jupyter AI.
+    """
+
+    name = "dataanalysis"
+    description = "Data exploration, statistical analysis, visualization, and ML insights"
+    version = "1.0.0"
+
+    # Phase 3: Framework file operations capability (read, write, edit, grep)
+    _file_ops = FileOperationsCapability()
+
+    @classmethod
+    def get_tools(cls) -> List[str]:
+        """Get the list of tools for data analysis tasks.
+
+        Phase 3: Uses framework FileOperationsCapability for common file operations
+        to reduce code duplication and maintain consistency across verticals.
+
+        Uses canonical tool names from victor.tools.tool_names.
+        """
+        from victor.tools.tool_names import ToolNames
+
+        # Start with framework file operations (read, write, edit, grep)
+        tools = cls._file_ops.get_tool_list()
+
+        # Add data analysis-specific tools
+        tools.extend(
+            [
+                # Directory listing for data file exploration
+                ToolNames.LS,  # list_directory → ls
+                # Python/Shell execution for analysis
+                ToolNames.SHELL,  # bash → shell (for running Python scripts)
+                # Code generation and search
+                ToolNames.CODE_SEARCH,  # Semantic code search
+                ToolNames.OVERVIEW,  # codebase_overview → overview
+                ToolNames.GRAPH,  # Code graph analysis (PageRank, dependencies)
+                # Web for datasets and documentation
+                ToolNames.WEB_SEARCH,  # Web search (internet search)
+                ToolNames.WEB_FETCH,  # Fetch URL content
+            ]
+        )
+
+        return tools
+
+    @classmethod
+    def get_system_prompt(cls) -> str:
+        """Get the system prompt for data analysis tasks."""
+        return cls._get_system_prompt()
+
+    @classmethod
+    def get_stages(cls) -> Dict[str, StageDefinition]:
+        """Get Data Analysis-specific stage definitions.
+
+        Uses canonical tool names from victor.tools.tool_names.
+        """
+        from victor.tools.tool_names import ToolNames
+
+        return {
+            "INITIAL": StageDefinition(
+                name="INITIAL",
+                description="Understanding the data and analysis goals",
+                tools={ToolNames.READ, ToolNames.LS, ToolNames.OVERVIEW},
+                keywords=["what", "data", "analyze", "understand", "explore"],
+                next_stages={"DATA_LOADING", "EXPLORATION"},
+            ),
+            "DATA_LOADING": StageDefinition(
+                name="DATA_LOADING",
+                description="Loading and validating data files",
+                tools={ToolNames.READ, ToolNames.SHELL, ToolNames.WRITE},
+                keywords=["load", "import", "read", "open", "fetch"],
+                next_stages={"EXPLORATION", "CLEANING"},
+            ),
+            "EXPLORATION": StageDefinition(
+                name="EXPLORATION",
+                description="Exploratory data analysis and profiling",
+                tools={ToolNames.SHELL, ToolNames.READ, ToolNames.WRITE},
+                keywords=["explore", "profile", "describe", "summary", "statistics"],
+                next_stages={"CLEANING", "ANALYSIS"},
+            ),
+            "CLEANING": StageDefinition(
+                name="CLEANING",
+                description="Data cleaning and transformation",
+                tools={ToolNames.SHELL, ToolNames.WRITE, ToolNames.EDIT},
+                keywords=["clean", "transform", "fix", "handle", "remove"],
+                next_stages={"ANALYSIS", "EXPLORATION"},
+            ),
+            "ANALYSIS": StageDefinition(
+                name="ANALYSIS",
+                description="Statistical analysis and modeling",
+                tools={ToolNames.SHELL, ToolNames.WRITE, ToolNames.READ},
+                keywords=["analyze", "model", "correlate", "test", "predict"],
+                next_stages={"VISUALIZATION", "REPORTING"},
+            ),
+            "VISUALIZATION": StageDefinition(
+                name="VISUALIZATION",
+                description="Creating charts and visualizations",
+                tools={ToolNames.SHELL, ToolNames.WRITE},
+                keywords=["plot", "chart", "visualize", "graph", "figure"],
+                next_stages={"REPORTING", "ANALYSIS"},
+            ),
+            "REPORTING": StageDefinition(
+                name="REPORTING",
+                description="Generating insights and reports",
+                tools={ToolNames.WRITE, ToolNames.EDIT, ToolNames.READ},
+                keywords=["report", "summarize", "document", "present"],
+                next_stages={"COMPLETION"},
+            ),
+            "COMPLETION": StageDefinition(
+                name="COMPLETION",
+                description="Finalizing analysis deliverables",
+                tools={ToolNames.WRITE, ToolNames.READ},
+                keywords=["done", "complete", "finish", "final"],
+                next_stages=set(),
+            ),
+        }
+
+    @classmethod
+    def _get_system_prompt(cls) -> str:
+        return """You are a data analysis assistant specializing in exploration, statistics, and visualization.
+
+## Core Capabilities
+
+1. **Data Loading**: CSV, Excel, JSON, Parquet, SQL databases
+2. **Exploration**: Profiling, summary statistics, distribution analysis
+3. **Cleaning**: Missing values, outliers, type conversion, normalization
+4. **Analysis**: Correlation, regression, hypothesis testing, clustering
+5. **Visualization**: matplotlib, seaborn, plotly for charts and dashboards
+6. **ML**: scikit-learn for classification, regression, clustering
+
+## Analysis Workflow
+
+1. **LOAD**: Read data, check structure, identify types
+2. **EXPLORE**: Summary stats, distributions, missing values
+3. **CLEAN**: Handle nulls, outliers, type issues
+4. **ANALYZE**: Apply statistical methods, test hypotheses
+5. **VISUALIZE**: Create informative charts
+6. **REPORT**: Summarize insights with evidence
+
+## Code Standards
+
+- Always use pandas for data manipulation
+- Include comments explaining methodology
+- Handle missing data explicitly
+- Use descriptive variable names
+- Save intermediate results for reproducibility
+
+## Output Format
+
+When presenting analysis:
+1. Start with data overview (shape, types, missing)
+2. Show key statistics with context
+3. Include visualizations with captions
+4. State insights with supporting evidence
+5. Note limitations and assumptions
+6. Provide reproducible code
+
+## Privacy and Ethics
+
+- Never expose personally identifiable information (PII)
+- Anonymize sensitive columns before analysis
+- Note potential biases in data
+- Be transparent about limitations
+"""
+
+    # =========================================================================
+    # New Framework Integrations (Workflows, RL, Teams)
+    # =========================================================================