Pipeline System

class Pipeline:
    """
    Directed graph of workflow stages for building multi-step applications.
    
    A Pipeline represents a UI workflow of multiple linear or branching stages,
    where each stage returns a Panel object to render.
    
    Parameters:
    - stages: Dictionary mapping stage names to stage functions or objects
    - start: Name of the starting stage
    - debug: Whether to show debugging information
    - sizing_mode: How the pipeline should size itself
    - **params: Additional parameters
    
    Layout Components:
    - header: Pipeline header with navigation controls
    - title: Name of the current stage
    - error: Error display field
    - network: Network diagram of the pipeline graph
    - buttons: Navigation buttons and selectors
    - prev_button: Button to go to previous stage
    - prev_selector: Selector for branching previous stages
    - next_button: Button to go to next stage
    - next_selector: Selector for branching next stages
    """

class PipelineError:
    """
    Custom error type for displaying pipeline-specific error messages.
    
    Raised within pipeline stages to display custom error messages
    in the pipeline's error display area.
    """

import panel as pn

def stage1():
    return pn.Column(
        "# Stage 1: Data Input",
        pn.widgets.FileInput(accept='.csv'),
        pn.widgets.Button(name="Next", button_type="primary")
    )

def stage2():
    return pn.Column(
        "# Stage 2: Data Processing", 
        pn.pane.Markdown("Processing your data..."),
        pn.widgets.Button(name="Next", button_type="primary")
    )

def stage3():
    return pn.Column(
        "# Stage 3: Results",
        pn.pane.Markdown("Your results are ready!"),
        pn.widgets.Button(name="Download", button_type="success")
    )

# Create linear pipeline
pipeline = pn.pipeline.Pipeline({
    'input': stage1,
    'process': stage2, 
    'results': stage3
}, start='input')

pipeline.servable()

import panel as pn

def analysis_choice():
    return pn.Column(
        "# Choose Analysis Type",
        pn.widgets.RadioButtonGroup(
            name="Analysis", 
            options=['Statistical', 'Machine Learning', 'Visualization'],
            value='Statistical'
        )
    )

def statistical_analysis():
    return pn.pane.Markdown("# Statistical Analysis Results")

def ml_analysis():
    return pn.pane.Markdown("# Machine Learning Results")

def visualization():
    return pn.pane.Markdown("# Data Visualization")

# Create branching pipeline
pipeline = pn.pipeline.Pipeline({
    'choice': (analysis_choice, ['stats', 'ml', 'viz']),
    'stats': statistical_analysis,
    'ml': ml_analysis,
    'viz': visualization
}, start='choice')