"""
Pattern 160: Inductive Reasoning
Description:
Inductive reasoning generalizes from specific observations to broader patterns
or rules. It moves from particular instances to general conclusions, inferring
patterns that likely hold across similar cases.
Components:
- Example collection
- Pattern identification
- Rule generalization
- Confidence assessment
- Counter-example checking
Use Cases:
- Pattern discovery
- Rule learning
- Trend analysis
- Scientific generalization
- Predictive modeling
Benefits:
- Discovers patterns from data
- Enables prediction
- Learns from examples
- Practical applicability
Trade-offs:
- Conclusions not guaranteed
- Vulnerable to biased samples
- Requires sufficient examples
- May over-generalize
LangChain Implementation:
Uses ChatOpenAI for pattern recognition, rule extraction,
and validation through additional examples
"""
import os
from typing import List, Dict, Any, Optional
from dataclasses import dataclass, field
from dotenv import load_dotenv
from langchain_openai import ChatOpenAI
from langchain_core.prompts import ChatPromptTemplate
from langchain_core.output_parsers import StrOutputParser
import json
load_dotenv()
@dataclass
class Example:
"""Single example/observation"""
input: str
output: str
features: Dict[str, Any] = field(default_factory=dict)
metadata: Dict[str, Any] = field(default_factory=dict)
@dataclass
class Pattern:
"""Discovered pattern"""
description: str
rule: str
confidence: float # 0-1
supporting_examples: List[str] = field(default_factory=list)
counter_examples: List[str] = field(default_factory=list)
generality: float = 0.5 # How general vs specific
@dataclass
class InductiveAnalysis:
"""Result of inductive reasoning"""
examples: List[Example]
patterns: List[Pattern]
generalizations: List[str] = field(default_factory=list)
confidence: float = 0.0
reasoning_trace: List[str] = field(default_factory=list)
class InductiveReasoningAgent:
"""Agent that performs inductive reasoning"""
def __init__(self, model_name: str = "gpt-3.5-turbo"):
"""
Initialize inductive reasoning agent
Args:
model_name: LLM model to use
"""
self.llm = ChatOpenAI(model=model_name, temperature=0.7)
self.validator_llm = ChatOpenAI(model=model_name, temperature=0)
def reason(self, examples: List[Example],
domain: Optional[str] = None,
min_confidence: float = 0.5) -> InductiveAnalysis:
"""
Perform inductive reasoning on examples
Args:
examples: List of examples to learn from
domain: Optional domain context
min_confidence: Minimum confidence threshold
Returns:
InductiveAnalysis with discovered patterns
"""
reasoning_trace = []
# Step 1: Identify patterns
reasoning_trace.append(f"Analyzing {len(examples)} examples...")
patterns = self._identify_patterns(examples, domain)
reasoning_trace.append(f"Identified {len(patterns)} potential patterns")
# Step 2: Validate patterns
reasoning_trace.append("Validating patterns against examples...")
for pattern in patterns:
self._validate_pattern(pattern, examples)
# Step 3: Filter by confidence
patterns = [p for p in patterns if p.confidence >= min_confidence]
reasoning_trace.append(f"Retained {len(patterns)} high-confidence patterns")
# Step 4: Generate generalizations
generalizations = self._generate_generalizations(patterns, examples)
reasoning_trace.append(f"Generated {len(generalizations)} generalizations")
# Step 5: Calculate overall confidence
if patterns:
avg_confidence = sum(p.confidence for p in patterns) / len(patterns)
else:
avg_confidence = 0.0
return InductiveAnalysis(
examples=examples,
patterns=patterns,
generalizations=generalizations,
confidence=avg_confidence,
reasoning_trace=reasoning_trace
)
def _identify_patterns(self, examples: List[Example],
domain: Optional[str]) -> List[Pattern]:
"""Identify patterns from examples"""
# Format examples
examples_text = "\n".join([
f"Input: {ex.input} → Output: {ex.output}"
for ex in examples
])
# Create pattern identification prompt
prompt = ChatPromptTemplate.from_messages([
("system", """You are an expert at identifying patterns from examples.
Analyze the given input-output examples and identify patterns or rules.
{domain_context}
For each pattern, provide:
1. Clear description
2. General rule
3. How general vs specific it is (0-1)
Return as JSON array: [{{"description": "...", "rule": "...", "generality": 0.0-1.0}}]"""),
("user", "Examples:\n{examples}\n\nIdentify patterns:")
])
chain = prompt | self.llm | StrOutputParser()
domain_context = f"Domain: {domain}" if domain else "Consider general patterns."
response = chain.invoke({
"examples": examples_text,
"domain_context": domain_context
})
# Parse patterns
patterns = self._parse_patterns(response)
return patterns
def _parse_patterns(self, response: str) -> List[Pattern]:
"""Parse patterns from LLM response"""
try:
# Try to extract JSON
start = response.find('[')
end = response.rfind(']') + 1
if start >= 0 and end > start:
json_str = response[start:end]
data = json.loads(json_str)
patterns = []
for item in data:
pattern = Pattern(
description=item.get('description', ''),
rule=item.get('rule', ''),
confidence=0.5, # Initial, will validate
generality=float(item.get('generality', 0.5))
)
patterns.append(pattern)
return patterns
except Exception as e:
print(f"Error parsing patterns: {e}")
# Fallback parsing
patterns = []
lines = response.split('\n')
current_pattern = {}
for line in lines:
line = line.strip()
if line.startswith(('1.', '2.', '3.', '4.', '5.')):
if current_pattern.get('description'):
patterns.append(Pattern(
description=current_pattern.get('description', ''),
rule=current_pattern.get('rule', ''),
confidence=0.5,
generality=0.5
))
current_pattern = {'description': line[2:].strip()}
elif 'rule:' in line.lower():
current_pattern['rule'] = line.split(':', 1)[1].strip()
# Add last pattern
if current_pattern.get('description'):
patterns.append(Pattern(
description=current_pattern.get('description', ''),
rule=current_pattern.get('rule', ''),
confidence=0.5,
generality=0.5
))
return patterns[:5] # Limit to 5
def _validate_pattern(self, pattern: Pattern, examples: List[Example]):
"""Validate pattern against examples"""
supporting = 0
contradicting = 0
for example in examples:
# Check if pattern applies to this example
prompt = ChatPromptTemplate.from_messages([
("system", "Determine if the pattern/rule applies to this example. Respond with 'yes' or 'no'."),
("user", """Pattern: {pattern}
Rule: {rule}
Example - Input: {input}, Output: {output}
Does the pattern apply?""")
])
chain = prompt | self.validator_llm | StrOutputParser()
response = chain.invoke({
"pattern": pattern.description,
"rule": pattern.rule,
"input": example.input,
"output": example.output
})
if 'yes' in response.lower():
supporting += 1
pattern.supporting_examples.append(f"{example.input} → {example.output}")
else:
contradicting += 1
pattern.counter_examples.append(f"{example.input} → {example.output}")
# Calculate confidence
total = supporting + contradicting
if total > 0:
pattern.confidence = supporting / total
else:
pattern.confidence = 0.0
def _generate_generalizations(self, patterns: List[Pattern],
examples: List[Example]) -> List[str]:
"""Generate general rules from patterns"""
if not patterns:
return []
# Format patterns
patterns_text = "\n".join([
f"- {p.description} (confidence: {p.confidence:.0%})"
for p in patterns
])
# Create generalization prompt
prompt = ChatPromptTemplate.from_messages([
("system", "Synthesize the patterns into broader generalizations or principles."),
("user", """Patterns discovered:
{patterns}
Generate 2-3 general rules or principles that capture these patterns:""")
])
chain = prompt | self.llm | StrOutputParser()
response = chain.invoke({"patterns": patterns_text})
# Parse generalizations
generalizations = []
for line in response.split('\n'):
line = line.strip()
if line and (line[0].isdigit() or line.startswith('-')):
clean_line = line.lstrip('0123456789.-) ')
if clean_line:
generalizations.append(clean_line)
return generalizations
def predict(self, analysis: InductiveAnalysis, new_input: str) -> Dict[str, Any]:
"""Use learned patterns to make a prediction"""
# Format patterns
patterns_text = "\n".join([
f"- {p.rule} (confidence: {p.confidence:.0%})"
for p in analysis.patterns
])
prompt = ChatPromptTemplate.from_messages([
("system", """Use the learned patterns to predict the output for the new input.
Explain which pattern(s) you're applying."""),
("user", """Learned patterns:
{patterns}
New input: {input}
Predict the output and explain your reasoning:""")
])
chain = prompt | self.llm | StrOutputParser()
response = chain.invoke({
"patterns": patterns_text,
"input": new_input
})
return {
"input": new_input,
"prediction": response,
"based_on_patterns": len(analysis.patterns),
"confidence": analysis.confidence
}
def demonstrate_inductive_reasoning():
"""Demonstrate inductive reasoning pattern"""
print("=" * 80)
print("INDUCTIVE REASONING PATTERN DEMONSTRATION")
print("=" * 80)
agent = InductiveReasoningAgent()
# Example 1: Number patterns
print("\n" + "="*80)
print("EXAMPLE 1: Number Pattern Discovery")
print("="*80)
examples = [
Example(input="2", output="4"),
Example(input="3", output="9"),
Example(input="4", output="16"),
Example(input="5", output="25"),
Example(input="6", output="36"),
]
analysis = agent.reason(examples, domain="Mathematics")
print("Examples:")
for ex in examples:
print(f" {ex.input} → {ex.output}")
print(f"\nDiscovered Patterns:")
for i, pattern in enumerate(analysis.patterns, 1):
print(f"\n{i}. {pattern.description}")
print(f" Rule: {pattern.rule}")
print(f" Confidence: {pattern.confidence:.0%}")
print(f" Generality: {pattern.generality:.1f}")
# Test prediction
if analysis.patterns:
prediction = agent.predict(analysis, "7")
print(f"\nPrediction for '7': {prediction['prediction'][:200]}...")
# Example 2: Text transformation
print("\n" + "="*80)
print("EXAMPLE 2: Text Transformation Pattern")
print("="*80)
examples = [
Example(input="cat", output="cats"),
Example(input="dog", output="dogs"),
Example(input="house", output="houses"),
Example(input="car", output="cars"),
Example(input="book", output="books"),
]
analysis = agent.reason(examples, domain="English grammar")
print("Examples:")
for ex in examples:
print(f" {ex.input} → {ex.output}")
print(f"\nPattern Analysis:")
for pattern in analysis.patterns:
print(f"\n- {pattern.description}")
print(f" Confidence: {pattern.confidence:.0%}")
if pattern.supporting_examples:
print(f" Supporting: {len(pattern.supporting_examples)} examples")
print(f"\nGeneralizations:")
for gen in analysis.generalizations:
print(f" - {gen}")
# Example 3: Category classification
print("\n" + "="*80)
print("EXAMPLE 3: Classification Pattern")
print("="*80)
examples = [
Example(input="apple", output="fruit"),
Example(input="banana", output="fruit"),
Example(input="carrot", output="vegetable"),
Example(input="broccoli", output="vegetable"),
Example(input="orange", output="fruit"),
Example(input="spinach", output="vegetable"),
]
analysis = agent.reason(examples, domain="Food classification")
print("Examples:")
for ex in examples:
print(f" {ex.input} → {ex.output}")
print(f"\nDiscovered Rules:")
for pattern in analysis.patterns:
print(f"\n- {pattern.rule}")
print(f" Confidence: {pattern.confidence:.0%}")
# Test with new input
prediction = agent.predict(analysis, "tomato")
print(f"\nPrediction for 'tomato':")
print(f" {prediction['prediction'][:300]}...")
# Example 4: Sequence patterns
print("\n" + "="*80)
print("EXAMPLE 4: Sequence Pattern")
print("="*80)
examples = [
Example(input="position 1", output="2"),
Example(input="position 2", output="4"),
Example(input="position 3", output="6"),
Example(input="position 4", output="8"),
Example(input="position 5", output="10"),
]
analysis = agent.reason(examples, domain="Sequences")
print("Examples:")
for ex in examples:
print(f" {ex.input} → {ex.output}")
print(f"\nPattern:")
if analysis.patterns:
pattern = analysis.patterns[0]
print(f" {pattern.description}")
print(f" Rule: {pattern.rule}")
print(f" Confidence: {pattern.confidence:.0%}")
# Example 5: Behavior pattern
print("\n" + "="*80)
print("EXAMPLE 5: Behavioral Pattern")
print("="*80)
examples = [
Example(input="Monday morning", output="heavy traffic"),
Example(input="Tuesday morning", output="heavy traffic"),
Example(input="Wednesday morning", output="heavy traffic"),
Example(input="Saturday morning", output="light traffic"),
Example(input="Sunday morning", output="light traffic"),
]
analysis = agent.reason(examples, domain="Traffic patterns")
print("Examples:")
for ex in examples:
print(f" {ex.input} → {ex.output}")
print(f"\nInduced Patterns:")
for i, pattern in enumerate(analysis.patterns, 1):
print(f"\n{i}. {pattern.description}")
print(f" Confidence: {pattern.confidence:.0%}")
print(f"\nGeneralizations:")
for gen in analysis.generalizations:
print(f" - {gen}")
# Prediction
prediction = agent.predict(analysis, "Thursday morning")
print(f"\nPrediction for 'Thursday morning':")
print(f" {prediction['prediction'][:200]}...")
# Summary
print("\n" + "="*80)
print("SUMMARY: Inductive Reasoning Best Practices")
print("="*80)
print("""
1. EXAMPLE COLLECTION:
- Gather diverse, representative examples
- Ensure sufficient quantity
- Include edge cases
- Verify example quality
2. PATTERN IDENTIFICATION:
- Look for commonalities
- Consider multiple levels of abstraction
- Identify exceptions
- Test alternative patterns
3. VALIDATION:
- Check against all examples
- Count supporting vs contradicting evidence
- Calculate confidence scores
- Seek counter-examples
4. GENERALIZATION:
- Move from specific to general
- Define scope of applicability
- Acknowledge limitations
- Avoid over-generalization
5. CONFIDENCE ASSESSMENT:
- Based on supporting evidence ratio
- Consider sample size
- Account for example diversity
- Update with new evidence
6. APPLICATIONS:
- Pattern discovery in data
- Rule learning from examples
- Trend analysis
- Predictive modeling
- Scientific generalization
Benefits:
✓ Discovers patterns from data
✓ Learns from examples
✓ Enables predictions
✓ Practical and applicable
✓ Foundation for machine learning
Limitations:
- Conclusions not guaranteed
- Sensitive to sample bias
- May over-generalize
- Requires sufficient examples
- Exception handling needed
""")
if __name__ == "__main__":
demonstrate_inductive_reasoning()
Agentic Design Pattern: Inductive Reasoning
Pattern 160: Inductive Reasoning – Line-by-Line Explanation
Overview
This document provides a detailed line-by-line explanation of the Inductive Reasoning pattern implementation.
Pattern Description
Pattern 160: Inductive Reasoning
Description: Inductive reasoning generalizes from specific observations to broader patterns or rules. It moves from particular instances to general conclusions, inferring patterns that likely hold across similar cases.
Components: – Example collection – Pattern identification – Rule generalization – Confidence assessment – Counter-example checking
Use Cases: – Pattern discovery – Rule learning – Trend analysis – Scientific generalization – Predictive modeling
Benefits: – Discovers patterns from data – Enables prediction – Learns from examples – Practical applicability
Trade-offs: – Conclusions not guaranteed – Vulnerable to biased samples – Requires sufficient examples – May over-generalize
LangChain Implementation: Uses ChatOpenAI for pattern recognition, rule extraction, and validation through additional examples
File Statistics
- Total Lines: 558
- Classes: 4
- Functions: 1
- Constants: 0
Module Documentation
"""
Pattern 160: Inductive Reasoning
Description:
Inductive reasoning generalizes from specific observations to broader patterns
or rules. It moves from particular instances to general conclusions, inferring
patterns that likely hold across similar cases.
Components:
- Example collection
- Pattern identification
- Rule generalization
- Confidence assessment
- Counter-example checking
Use Cases:
- Pattern discovery
- Rule learning
- Trend analysis
- Scientific generalization
- Predictive modeling
Benefits:
- Discovers patterns from data
- Enables prediction
- Learns from examples
- Practical applicability
Trade-offs:
- Conclusions not guaranteed
- Vulnerable to biased samples
- Requires sufficient examples
- May over-generalize
LangChain Implementation:
Uses ChatOpenAI for pattern recognition, rule extraction,
and validation through additional examples
"""
Purpose: This module implements the Inductive Reasoning pattern, providing a structured approach to agent design and implementation.
Import Statements
Standard Library Imports
import dataclasses.dataclass
import dataclasses.field
import json
import langchain_core.output_parsers.StrOutputParser
import langchain_core.prompts.ChatPromptTemplate
import os
import typing.Any
import typing.Dict
import typing.List
import typing.Optional
Purpose: Core Python functionality:
– dataclasses.dataclass: Simplified class definitions with @dataclass
– dataclasses.field: Simplified class definitions with @dataclass
– json: JSON serialization/deserialization
– langchain_core.output_parsers.StrOutputParser: Regular expression operations
– langchain_core.prompts.ChatPromptTemplate: Regular expression operations
– os: Environment variable access and system operations
– typing.Any: Type hints for better code documentation (List, Dict, Any, Optional)
– typing.Dict: Type hints for better code documentation (List, Dict, Any, Optional)
– typing.List: Type hints for better code documentation (List, Dict, Any, Optional)
– typing.Optional: Type hints for better code documentation (List, Dict, Any, Optional)
LangChain Imports
from langchain_openai.ChatOpenAI
Purpose: LangChain framework components: – LLM Interface: Wrapper for OpenAI models – Chains: Composable building blocks for complex workflows
Other Imports
from dotenv.load_dotenv
Purpose: Additional dependencies for specialized functionality.
Classes
Class: Example
Lines 54-59
Purpose:
Single example/observation
Class: Pattern
Lines 63-70
Purpose:
Discovered pattern
Class: InductiveAnalysis
Lines 74-80
Purpose:
Result of inductive reasoning
Class: InductiveReasoningAgent
Lines 83-342
Purpose:
Agent that performs inductive reasoning
Methods:
__init__(self, model_name)
Lines 86-94
Purpose:
Initialize inductive reasoning agent
Args:
model_name: LLM model to use
Parameters:
– model_name: Language model instance
reason(self, examples, domain, min_confidence)
Lines 96-142
Purpose:
Perform inductive reasoning on examples
Args:
examples: List of examples to learn from
domain: Optional domain context
min_confidence: Minimum confidence threshold
Returns:
InductiveAnalysis with discovered patterns
Parameters:
– examples: Examples
– domain: Domain
– min_confidence: Min confidence
Returns: InductiveAnalysis
_identify_patterns(self, examples, domain)
Lines 144-180
Purpose:
Identify patterns from examples
Parameters:
– examples: Examples
– domain: Domain
Returns: <ast.Subscript object at 0x000001A06D684350>
_parse_patterns(self, response)
Lines 182-233
Purpose:
Parse patterns from LLM response
Parameters:
– response: Response
Returns: <ast.Subscript object at 0x000001A06D615FD0>
_validate_pattern(self, pattern, examples)
Lines 235-273
Purpose:
Validate pattern against examples
Parameters:
– pattern: Pattern
– examples: Examples
_generate_generalizations(self, patterns, examples)
Lines 275-309
Purpose:
Generate general rules from patterns
Parameters:
– patterns: Patterns
– examples: Examples
Returns: <ast.Subscript object at 0x000001A06D7CC990>
predict(self, analysis, new_input)
Lines 311-342
Purpose:
Use learned patterns to make a prediction
Parameters:
– analysis: Analysis
– new_input: Input data
Returns: <ast.Subscript object at 0x000001A06D8550D0>
Functions
demonstrate_inductive_reasoning()
Lines 345-553
Purpose:
Demonstrate inductive reasoning pattern
Key Concepts Summary
Pattern Overview
The Inductive Reasoning pattern provides a structured approach to building AI agents with specific capabilities and behaviors. This implementation leverages LangChain’s composable components to create a robust and maintainable solution.
Core Components
- Agent Class: Main implementation of the pattern logic
- Prompt Engineering: Carefully crafted prompts to guide LLM behavior
- Tool Integration: Optional tools for extended functionality
- Memory Management: State handling across interactions
- Output Parsing: Structured response processing
When to Use
✅ Good for: – Scenarios requiring this specific pattern’s strengths – Applications needing structured agent behavior – Systems where explainability is important – Production environments requiring reliability
❌ Not ideal for: – Simple single-prompt tasks – Scenarios requiring different pattern characteristics – Resource-constrained environments (if computationally intensive)
Best Practices Demonstrated
1. Environment Configuration
- ✅ Use
.envfiles for API keys - ✅ Never hardcode credentials
- ✅ Validate environment variables on startup
2. Error Handling
- ✅ Graceful degradation on failures
- ✅ Informative error messages
- ✅ Logging for debugging
3. Prompt Engineering
- ✅ Clear, specific instructions
- ✅ Structured output formats
- ✅ Few-shot examples when beneficial
4. Code Organization
- ✅ Separation of concerns (class-based design)
- ✅ Type hints for clarity
- ✅ Docstrings for documentation
5. Performance Optimization
- ✅ Appropriate temperature settings
- ✅ Token budget management
- ✅ Caching where applicable
LangChain Components Used
- ChatOpenAI: OpenAI model interface
- Chains: Sequential component composition
Why LangChain? – Composable abstractions for LLM applications – Rich ecosystem of integrations – Production-ready patterns – Active community and development
Execution Flow
1. Initialize agent with configuration
↓
2. Load environment and dependencies
↓
3. Prepare prompt with input data
↓
4. Invoke LLM through LangChain
↓
5. Parse and validate response
↓
6. Return formatted result
Key Decision Points: – Model selection (temperature, model type) – Prompt structure (few-shot vs zero-shot) – Error handling strategy – Output format validation
Performance Considerations
Computational Cost
- LLM Calls: Primary cost factor (based on tokens)
- Latency: Depends on model and input size
- Memory: Minimal for most patterns
Optimization Strategies
- Prompt Optimization: Minimize tokens while maintaining clarity
- Caching: Cache repeated queries/responses
- Model Selection: Use smaller models when appropriate
- Batch Processing: Group similar requests
- Streaming: For long responses, stream tokens
Cost Estimates (Approximate)
- GPT-3.5-turbo: ~$0.002 per request
- GPT-4: ~$0.03 per request
- GPT-4-turbo: ~$0.01 per request
Actual costs vary based on input/output token counts
Common Pitfalls to Avoid
1. Prompt Design Issues
❌ Vague or ambiguous instructions
✅ Clear, specific, testable prompts
2. Error Handling
❌ Ignoring API errors or rate limits
✅ Implement retry logic and graceful degradation
3. Token Management
❌ Exceeding model token limits
✅ Truncate inputs, chunk processing
4. Security
❌ Exposing API keys in code
✅ Use environment variables
5. Testing
❌ Only testing happy paths
✅ Test edge cases, failures, malformed inputs
6. Cost Management
❌ Unmonitored API usage
✅ Implement budgets, alerts, and monitoring
Further Improvements
Production Enhancements
- Monitoring: Add logging, metrics, tracing
- Validation: Input/output schema validation
- Rate Limiting: Implement request throttling
- Caching: Redis or in-memory cache for responses
- Async Support: Non-blocking operations for scalability
Feature Extensions
- Multi-Model Support: Switch between providers
- Fallback Models: Graceful degradation strategy
- Custom Tools: Domain-specific function calling
- Advanced Memory: Vector stores for long-term memory
- Evaluation: Automated testing and benchmarking
Code Quality
- Testing: Unit tests, integration tests
- Documentation: API docs, usage examples
- Type Safety: Stricter type hints
- Configuration: Externalized config files
Related Patterns
- Pattern 159
- Pattern 161
Pattern Families: – Reasoning Patterns (001-011) – Multi-Agent Patterns (012-019) – Tool Use Patterns (020-025) – Memory Patterns (026-031) – And many more…
Summary
The Inductive Reasoning pattern provides a powerful approach to building AI agents with specific capabilities. This implementation demonstrates:
✅ Key Strengths: – Clean, maintainable code structure – LangChain integration for composability – Type hints and documentation – Production-ready patterns
🎯 Best Use Cases: – Applications requiring this pattern’s characteristics – Systems needing structured agent behavior – Production environments requiring reliability
📚 Learning Resources: – LangChain Documentation: https://python.langchain.com/ – Pattern Documentation: See related patterns – OpenAI API Reference: https://platform.openai.com/docs/
File: 160_inductive_reasoning.py
Total Lines: 558
Pattern Category: AI Agent Design Pattern #160
Generated from actual implementation code
For educational and reference purposes