AgentFlow v0.1.1 - Workflow Transform Functions Update

AgentFlow v0.1.1 Release Notes

🚀 Major Updates

1. Workflow Transform Functions

Core Updates

• ✨ Updated transform functions to accept both step and context parameters
• 🔧 Added feature engineering transform with StandardScaler support
• 🛠️ Added outlier removal transform with IsolationForest support

Testing Improvements

• ✅ Added comprehensive unit tests for transform functions
• 📊 Added performance tests for workflow execution
• 🔍 Enhanced test coverage and error handling tests

2. Documentation Updates

• 📚 Updated all README files with latest version information
• 💡 Added detailed examples for transform functions
• 📖 Improved API documentation and usage guides

3. Error Handling

• ⚡ Enhanced validation for transform function parameters
• 🔔 Improved error messages and debugging information
• 🔄 Added retry policies for workflow steps

🔄 Breaking Changes

Transform functions now require both step and context parameters:

async def your_transform(
    step: WorkflowStep,
    context: Dict[str, Any]
) -> Dict[str, Any]:
    """Transform function with new parameter structure.
    
    Args:
        step: The workflow step being executed
        context: The execution context containing the data
        
    Returns:
        Dict containing the transformed data
    """
    return {"data": transformed_data}

📋 Migration Guide

1. Update your transform functions:

# Before
async def old_transform(data: Dict[str, Any]) -> Dict[str, Any]:
    return {"data": process(data)}

# After
async def new_transform(step: WorkflowStep, context: Dict[str, Any]) -> Dict[str, Any]:
    data = context["data"]
    return {"data": process(data)}

2. Update your workflow configurations:

# Before
workflow_config = WorkflowConfig(
    steps=[
        WorkflowStep(
            id="step-1",
            type=WorkflowStepType.TRANSFORM,
            config={"strategy": "standard"}
        )
    ]
)

# After
workflow_config = WorkflowConfig(
    steps=[
        WorkflowStep(
            id="step-1",
            type=WorkflowStepType.TRANSFORM,
            config=StepConfig(
                strategy="standard",
                params={"execute": new_transform}
            )
        )
    ]
)

🎯 Examples

Feature Engineering Transform

async def feature_engineering_transform(step: WorkflowStep, context: Dict[str, Any]) -> Dict[str, Any]:
    data = context["data"]
    scaler = StandardScaler(
        with_mean=step.config.params["with_mean"],
        with_std=step.config.params["with_std"]
    )
    transformed_data = scaler.fit_transform(data)
    return {"data": transformed_data}

Outlier Removal Transform

async def outlier_removal_transform(step: WorkflowStep, context: Dict[str, Any]) -> Dict[str, Any]:
    data = context["data"]
    iso_forest = IsolationForest(
        contamination=step.config.params["threshold"],
        random_state=42
    )
    predictions = iso_forest.fit_predict(data)
    filtered_data = data[predictions == 1]
    return {"data": filtered_data}

📝 Additional Notes

• All tests have been updated to reflect the new parameter structure
• Documentation has been updated with new examples
• Error messages now provide more context about parameter requirements