Tools | WQ

Tool Categories

Data Tools

load_csv, wrangle_data, show_info, describe_stats

Model Tools

split_features, fit_model, predict, calculate_mae

Viz Tools

create_histogram, scatter, mapbox, heatmap

Utility Tools

show_capabilities, run_lesson, export_data

Project Tools (2-3 per project)

Project	Tools
Proj2	`load_proj2_data`, `analyze_real_estate`, `predict_price`
Proj3	`load_proj3_data`, `analyze_air_quality`, `forecast_pm25`
Proj4	`load_proj4_data`, `classify_damage`, `evaluate_classifier`
Proj5	`load_proj5_data`, `detect_bankruptcy`, `handle_imbalance`
Proj6	`load_proj6_data`, `cluster_consumers`, `visualize_clusters`
Proj7	`load_proj7_data`, `run_chi_square`, `run_ab_test`
Proj8	`load_proj8_data`, `fit_garch`, `forecast_volatility`

Cross-Project Tools

Tool	Description
`compare_techniques`	Compare model performance across different techniques
`apply_model_to_data`	Apply a model from one project to data from another
`find_similar_datasets`	Find datasets with similar characteristics
`suggest_technique`	Get technique recommendations based on data type

Textbook Tools

Tool	Description
`textbook_search`	Search across 22 textbook chapters
`textbook_explain`	Get explanation of a concept with examples
`textbook_chapters`	List all available textbook chapters

Utility Tools

Tool	Description
`show_capabilities`	List all available tools with descriptions
`get_project_info`	Get metadata for any project
`export_data`	Export results to CSV/JSON format
`visualize_result`	Create visualization from model output
`run_lesson`	Execute a complete lesson workflow

Usage Example

from unified_agent import UnifiedDataClient, ToolRegistry # Initialize client = UnifiedDataClient() tools = ToolRegistry(client) # Load project data data = tools.get_tool('load_proj2_data')({'city': 'buenos_aires'}) # Run analysis result = tools.get_tool('analyze_real_estate')({'version': 1}) # Cross-project comparison comparison = tools.get_tool('compare_techniques')({ 'techniques': ['random_forest', 'gradient_boosting'], 'project': 'proj5' })