7.4 KiB
Workflow Statistics
ComfyFlow Insight
Technical Documentation
Overview
The Workflow Statistics Engine is responsible for converting raw workflow data into measurable engineering metrics.
While the Architecture Analyzer focuses on workflow structure, the Workflow Statistics Engine focuses on numerical measurements.
Its purpose is to answer questions such as:
- How large is the workflow?
- How complex is the workflow?
- How dense are the node connections?
- How interconnected is the workflow?
- How scalable is the design?
- How does this workflow compare to others?
The Statistics Engine provides objective measurements that can be used for workflow auditing, optimization, documentation, and engineering reviews.
Why Workflow Statistics Matter
Many workflow evaluations rely on visual inspection.
Visual inspection becomes unreliable when workflows exceed:
50+ Nodes
100+ Links
Multiple Branches
Workflow Statistics provides quantifiable metrics that remove guesswork.
Instead of saying:
"This workflow looks complicated."
You can say:
Node Count : 128
Link Count : 194
Complexity Score : 86
Workflow Scale : Advanced
Statistics Engine Objectives
The engine performs the following tasks:
- Count Nodes
- Count Links
- Calculate Workflow Density
- Calculate Average Connections
- Calculate Complexity Score
- Determine Workflow Scale
- Generate Statistical Summary
Node Count
Definition
Node Count represents the total number of nodes present in the workflow.
Every node contributes to workflow complexity.
Example
LoadImage
ImageScale
KSampler
SaveImage
Node Count:
4
Interpretation
1 – 25 Small Workflow
26 – 75 Medium Workflow
76 – 150 Large Workflow
150+ Enterprise Workflow
Why It Matters
Higher node counts generally indicate:
- More functionality
- Greater flexibility
- Increased maintenance requirements
- Higher debugging effort
Link Count
Definition
Link Count represents the total number of node connections.
Links define workflow relationships.
Example
LoadImage
↓
ImageScale
↓
KSampler
↓
SaveImage
Link Count:
3
Interpretation
High link counts indicate:
- Increased workflow interaction
- Strong dependency chains
- Greater execution complexity
Workflow Density
Definition
Workflow Density measures how interconnected the workflow is.
Density compares the number of links to the number of nodes.
Formula
Workflow Density = Link Count ÷ Node Count
Example
Workflow:
Nodes = 100
Links = 150
Density:
150 ÷ 100 = 1.5
Interpretation
0.0 – 1.0 Sparse
1.0 – 2.0 Moderate
2.0 – 3.0 Dense
3.0+ Highly Dense
Benefits
Density reveals:
- Workflow interconnectedness
- Dependency concentration
- Potential maintenance complexity
Average Connections
Definition
Average Connections measures how many relationships exist per node.
Formula
Average Connections =
Total Links ÷ Total Nodes
Example
Nodes = 120
Links = 180
Average Connections:
180 ÷ 120 = 1.5
Why It Matters
Higher values indicate:
- Increased node interaction
- More dependencies
- Higher workflow coupling
Complexity Score
Definition
Complexity Score is the primary engineering metric used to estimate workflow complexity.
It combines multiple measurements into a single value.
Inputs
The Complexity Score considers:
Node Count
Link Count
Workflow Depth
Branch Count
Dependency Complexity
Example
Workflow:
Nodes = 120
Links = 180
Depth = 28
Branches = 7
Result:
Complexity Score = 87
Interpretation
0 – 25 Simple
26 – 50 Moderate
51 – 75 Advanced
76 – 100 Enterprise
Benefits
Complexity Score provides:
- Workflow benchmarking
- Engineering assessment
- Portfolio metrics
- Workflow comparison
Workflow Scale
Definition
Workflow Scale categorizes workflow size.
Small
Characteristics:
Minimal Processing
Few Dependencies
Simple Design
Typical Size:
1 – 25 Nodes
Medium
Characteristics:
Multiple Processing Stages
Moderate Dependencies
Controlled Branching
Typical Size:
25 – 75 Nodes
Large
Characteristics:
Advanced Processing
Complex Relationships
Significant Branching
Typical Size:
75 – 150 Nodes
Enterprise
Characteristics:
Production Scale
Large Dependency Chains
Extensive Branching
Team Collaboration
Typical Size:
150+ Nodes
Statistical Summary
The Statistics Engine generates a summary report.
Example:
Workflow Statistics
Node Count : 120
Link Count : 180
Density : 1.50
Average Connections : 1.50
Complexity Score : 87
Workflow Scale : Advanced
Engineering Interpretation
Low Node Count + High Density
Possible Meaning:
Highly Optimized Workflow
Strong Node Reuse
Efficient Design
High Node Count + Low Density
Possible Meaning:
Large Workflow
Loose Coupling
Potential Redundancy
High Complexity Score
Possible Meaning:
Advanced Processing
Multiple Branches
Large Dependency Chains
Recommended Action:
Review Architecture
Review Branch Analysis
Review Dependency Graph
Common Workflow Categories
Basic Image Generation
Typical Statistics:
Nodes : 15–30
Links : 20–40
Complexity : Low
Advanced Image Generation
Typical Statistics:
Nodes : 50–120
Links : 80–200
Complexity : Medium
Video Generation Workflow
Typical Statistics:
Nodes : 80–200
Links : 120–300
Complexity : High
Production Workflow
Typical Statistics:
Nodes : 150+
Links : 250+
Complexity : Enterprise
Relationship With Other Engines
Workflow Statistics works alongside:
Architecture Analyzer
Branch Analyzer
Dependency Graph
Workflow Advisor
Statistics provide the numerical foundation used by higher-level analysis engines.
Best Practices
Monitor Complexity Growth
As workflows evolve:
- Track Node Count
- Track Link Count
- Monitor Density
- Monitor Complexity Score
Compare Workflow Versions
Statistics allow direct comparison between workflow revisions.
Example:
Version 1
Nodes = 80
Complexity = 62
Version 2
Nodes = 105
Complexity = 79
This immediately reveals complexity growth.
Summary
The Workflow Statistics Engine transforms raw workflow structures into measurable engineering metrics.
It provides:
- Node Count
- Link Count
- Workflow Density
- Average Connections
- Complexity Score
- Workflow Scale
These metrics help users understand workflow size, complexity, maintainability, and scalability.
They also provide the numerical foundation used by Workflow Advisor and Workflow Repair systems.
Next Document
Continue with:
docs/branch-analyzer.md
This document explains branch detection, branching complexity, parallel execution paths, workflow splitting, and branch risk analysis.