Pattern Analysis Example

This example demonstrates Kwasa-Kwasa’s powerful pattern analysis capabilities, showing how to extract meaning from text through various pattern recognition techniques, including visual patterns, letter frequencies, and orthographic analysis.

Source Code

// Example of pattern-based meaning extraction using Turbulance

// Some example text with potentially interesting patterns
item text = "The quick brown fox jumps over the lazy dog. How vexingly quick daft zebras jump!"

// Function to analyze letter frequency
funxn letter_frequency(text):
    item frequencies = {}
    item total = 0
    
    within text as characters:
        given character.is_alpha():
            item char_lower = character.lower()
            
            given char_lower in frequencies:
                frequencies[char_lower] = frequencies[char_lower] + 1
            given otherwise:
                frequencies[char_lower] = 1
                
            total = total + 1
    
    // Convert to percentages
    for each letter in frequencies:
        frequencies[letter] = frequencies[letter] / total
    
    return frequencies

// Calculate Shannon entropy of text
funxn calculate_entropy(text):
    item frequencies = letter_frequency(text)
    item entropy = 0.0
    
    for each letter, freq in frequencies:
        given freq > 0:
            entropy = entropy - freq * log2(freq)
    
    return entropy

// Detect recurring visual patterns
funxn detect_visual_patterns(text, pattern_length=3):
    // Map letters to shape classes
    item shape_classes = {
        'a': 0, 'c': 0, 'e': 0, 'o': 0, 's': 0,  // round shapes
        'i': 1, 'l': 1, 'j': 1, 'f': 1, 't': 1,  // vertical strokes
        'm': 2, 'n': 2, 'h': 2, 'u': 2,          // arch shapes
        'v': 3, 'w': 3, 'x': 3, 'y': 3, 'z': 3,  // angled shapes
        'b': 4, 'd': 4, 'p': 4, 'q': 4, 'g': 4   // circles with stems
    }
    // ... rest of the function

Code Explanation

1. Letter Frequency Analysis

funxn letter_frequency(text):
    item frequencies = {}
    item total = 0
    
    within text as characters:
        given character.is_alpha():
            // ... character counting logic

This function:

• Iterates through text characters
• Counts frequency of each letter
• Converts counts to percentages
• Useful for:
  • Statistical analysis
  • Language identification
  • Cryptographic analysis

2. Information Theory Analysis

funxn calculate_entropy(text):
    item frequencies = letter_frequency(text)
    item entropy = 0.0
    // ... entropy calculation

Calculates Shannon entropy of text:

• Measures information content
• Higher entropy = more randomness/information
• Lower entropy = more predictable patterns

3. Visual Pattern Recognition

funxn detect_visual_patterns(text, pattern_length=3):
    item shape_classes = {
        'a': 0, 'c': 0, 'e': 0, 'o': 0, 's': 0,  // round shapes
        'i': 1, 'l': 1, 'j': 1, 'f': 1, 't': 1,  // vertical strokes
        // ... more shape classes
    }

Features:

• Groups letters by visual shape
• Detects recurring shape patterns
• Useful for:
  • Typography analysis
  • Visual rhythm detection
  • Pattern-based meaning extraction

4. Orthographic Analysis

funxn visual_rhythm(text):
    // ... rhythm analysis code

funxn orthographic_density(text, width=40):
    // ... density mapping code

These functions analyze:

• Visual rhythm of text
• Character density patterns
• Typographic weight distribution
• Spatial relationships between characters

5. Statistical Pattern Analysis

funxn unusual_combinations(text, ngram_size=2):
    // ... n-gram analysis code

Identifies:

• Unusual letter combinations
• Deviations from expected frequencies
• Statistical anomalies in text

Running the Example

1. Save the code in a file with .turb extension
2. Run using the Kwasa-Kwasa interpreter:

   kwasa run pattern_analysis.turb
   

Expected Output

Analyzing text: The quick brown fox jumps over the lazy dog...

Letter frequencies:
  a: 0.0476
  b: 0.0238
  c: 0.0238
  ...

Text entropy: 4.2876 bits

Recurring visual patterns (by shape class):
  Pattern (0, 1, 2): ['ace', 'ane']
  Pattern (1, 0, 3): ['lex', 'laz']
  ...

Visual rhythm analysis (first 10 points):
  [0.4, 0.5, 0.7, 0.5, 0.2, 0.4, 0.5, 0.7, 0.4, 0.5]

Orthographic density map:
  ▁▂▃▅▂▁▃▄▂▁
  ▂▄▆▃▁▂▄▃▁▂
  ▁▃▂▁▄▅▂▁▃▄

Key Concepts Demonstrated

1. Statistical Analysis:
   • Letter frequency counting
   • Shannon entropy calculation
   • N-gram analysis
   • Statistical anomaly detection
2. Visual Analysis:
   • Shape-based pattern recognition
   • Visual rhythm detection
   • Orthographic density mapping
   • Typographic weight analysis
3. Pattern Recognition:
   • Recurring shape patterns
   • Letter combination analysis
   • Visual pattern matching
   • Multi-level pattern detection
4. Information Theory:
   • Entropy calculation
   • Information density analysis
   • Pattern frequency analysis
   • Statistical significance testing

Applications

1. Text Analysis:
   • Style analysis
   • Author identification
   • Language detection
   • Text complexity measurement
2. Visual Design:
   • Typography analysis
   • Layout optimization
   • Visual rhythm enhancement
   • Design pattern detection
3. Pattern Discovery:
   • Hidden pattern detection
   • Anomaly identification
   • Structure analysis
   • Pattern-based meaning extraction