Add Pattern Recognition & Visualization (Phase 2) #21

Copilot · 2025-12-01T09:10:53Z

Implements chart pattern detection with confidence scoring and visualization support via deepentropy/lwcharts.

New Packages

`src/numta/patterns/`

swing.py: Swing high/low detection (find_swing_highs, find_swing_lows, find_swing_points)
utils.py: Fibonacci retracement/extension, trendline fitting, confidence scoring
chart_patterns.py: Head & Shoulders, Double/Triple Top/Bottom, Triangles, Wedges, Flags, VCP
harmonic_patterns.py: Gartley, Butterfly, Bat, Crab patterns with standard Fibonacci ratios

`src/numta/viz/`

lwcharts_adapter.py: plot_chart(), plot_pattern(), plot_harmonic() with graceful degradation
pattern_markers.py: Annotation helpers for markers, trendlines, zones

Pandas Accessor

# Chart patterns
patterns = df.ta.find_patterns(pattern_type='double', order=5)

# Harmonic patterns  
harmonics = df.ta.find_harmonic_patterns(patterns=['gartley', 'bat'])

# Visualization
df.ta.plot(indicators={'SMA_20': df.ta.sma(20)}, patterns=patterns)

Optional Dependencies

viz = ["pandas>=1.3.0", "lwcharts>=0.1.0"]
patterns = ["pandas>=1.3.0"]
full = ["pandas>=1.3.0", "lwcharts>=0.1.0", "numba>=0.56.0"]

Tests

58 new tests in test_patterns.py and test_viz.py
All 302 tests pass

Original prompt

Phase 2: Pattern Recognition & Visualization

Overview

Implement chart pattern detection capabilities with visualization support using the deepentropy/lwcharts library (TradingView Lightweight Charts for Jupyter notebooks).

Deliverables

1. Chart Pattern Detection Module (src/numta/patterns/)

Create a new patterns subpackage:
src/numta/patterns/
├── __init__.py
├── chart_patterns.py        # Head & Shoulders, Double Top/Bottom, Triangles, Wedges, Flags, VCP
├── harmonic_patterns.py     # Gartley, Butterfly, Bat, Crab patterns
├── swing.py                 # Swing high/low detection
└── utils.py                 # Fibonacci levels, trendline fitting
Swing Detection (swing.py)
def find_swing_highs(high: np.ndarray, order: int = 5) -> np.ndarray
def find_swing_lows(low: np.ndarray, order: int = 5) -> np.ndarray
def find_swing_points(high: np.ndarray, low: np.ndarray, order: int = 5) -> tuple
Chart Patterns (chart_patterns.py)

Implement pattern dataclasses and detection functions:

HeadShouldersPattern + detect_head_shoulders(), detect_inverse_head_shoulders()

DoublePattern + detect_double_top(), detect_double_bottom()

TriplePattern + detect_triple_top(), detect_triple_bottom()

TrianglePattern + detect_triangle() (ascending, descending, symmetrical)

WedgePattern + detect_wedge() (rising, falling)

FlagPattern + detect_flag() (bull flag, bear flag, pennant)

VCPPattern + detect_vcp() (Volatility Contraction Pattern)

Each pattern should include: pattern_type, key points (indices + prices), trendlines/necklines, breakout levels, start/end indices, and confidence score (0-1).

Harmonic Patterns (harmonic_patterns.py)
@dataclass
class HarmonicPattern:
    pattern_type: str  # 'gartley', 'butterfly', 'bat', 'crab'
    X, A, B, C, D: tuple[int, float]  # Points with (index, price)
    direction: str  # 'bullish' or 'bearish'
    XA_retracement, AB_retracement, BC_retracement, CD_retracement: float
    prz: tuple[float, float]  # Potential Reversal Zone
    confidence: float

def detect_gartley(high, low, close, tolerance=0.02) -> list[HarmonicPattern]
def detect_butterfly(high, low, close, tolerance=0.02) -> list[HarmonicPattern]
def detect_bat(high, low, close, tolerance=0.02) -> list[HarmonicPattern]
def detect_crab(high, low, close, tolerance=0.02) -> list[HarmonicPattern]
def detect_harmonic_patterns(high, low, close, patterns=None, tolerance=0.02) -> list[HarmonicPattern]
Use standard Fibonacci ratios for each pattern type.

Utilities (utils.py)
def fibonacci_retracement(start: float, end: float) -> dict[str, float]
def fibonacci_extension(start: float, end: float, retracement: float) -> dict[str, float]
def fit_trendline(points: list[tuple[int, float]]) -> tuple[float, float]
def calculate_pattern_confidence(pattern_data: dict) -> float
2. Visualization Module (src/numta/viz/)
src/numta/viz/
├── __init__.py
├── lwcharts_adapter.py      # Integration with deepentropy/lwcharts
└── pattern_markers.py       # Pattern annotation helpers
def plot_chart(df, indicators=None, volume=True, height=400, width=800)
def plot_pattern(df, patterns, show_annotations=True, show_trendlines=True)
def plot_harmonic(df, patterns, show_fibonacci=True, show_prz=True)
3. Pandas Accessor Extension

Add to TAAccessor class in src/numta/pandas_ext.py:
def find_patterns(self, pattern_type='all', order=5, tolerance=0.03) -> list
def find_harmonic_patterns(self, patterns=None, tolerance=0.02) -> list[HarmonicPattern]
def plot(self, indicators=None, patterns=None, volume=True, height=400, width=800)
4. Update pyproject.toml
[project.optional-dependencies]
viz = ["pandas>=1.3.0", "lwcharts>=0.1.0"]
patterns = ["pandas>=1.3.0"]
full = ["pandas>=1.3.0", "lwcharts>=0.1.0", "numba>=0.56.0"]
5. Update src/numta/__init__.py

Export all pattern detection functions and dataclasses.

6. Tests

Create tests/test_patterns.py and tests/test_viz.py with comprehensive coverage.

7. Documentation

Update README.md with Pattern Recognition section including examples.

Acceptance Criteria

All chart patterns detected correctly with confidence scores

Harmonic patterns use correct Fibonacci ratios

Visualization works in Jupyter notebooks via lwcharts

Pandas .ta.find_patterns() and .ta.plot() methods work

Tests pass for all pattern types

Graceful degradation when lwcharts not installed

This pull request was created as a result of the following prompt from Copilot chat.

Phase 2: Pattern Recognition & Visualization

Overview

Implement chart pattern detection capabilities with visualization support using the deepentropy/lwcharts library (TradingView Lightweight Charts for Jupyter notebooks).

Deliverables

1. Chart Pattern Detection Module (src/numta/patterns/)

Create a new patterns subpackage:
src/numta/patterns/
├── __init__.py
├── chart_patterns.py        # Head & Shoulders, Double Top/Bottom, Triangles, Wedges, Flags, VCP
├── harmonic_patterns.py     # Gartley, Butterfly, Bat, Crab patterns
├── swing.py                 # Swing high/low detection
└── utils.py                 # Fibonacci levels, trendline fitting
Swing Detection (swing.py)
def find_swing_highs(high: np.ndarray, order: int = 5) -> np.ndarray
def find_swing_lows(low: np.ndarray, order: int = 5) -> np.ndarray
def find_swing_points(high: np.ndarray, low: np.ndarray, order: int = 5) -> tuple
Chart Patterns (chart_patterns.py)

Implement pattern dataclasses and detection functions:

HeadShouldersPattern + detect_head_shoulders(), detect_inverse_head_shoulders()

DoublePattern + detect_double_top(), detect_double_bottom()

TriplePattern + detect_triple_top(), detect_triple_bottom()

TrianglePattern + detect_triangle() (ascending, descending, symmetrical)

WedgePattern + detect_wedge() (rising, falling)

FlagPattern + detect_flag() (bull flag, bear flag, pennant)

VCPPattern + detect_vcp() (Volatility Contraction Pattern)

Each pattern should include: pattern_type, key points (indices + prices), trendlines/necklines, breakout levels, start/end indices, and confidence score (0-1).

Harmonic Patterns (harmonic_patterns.py)
@dataclass
class HarmonicPattern:
    pattern_type: str  # 'gartley', 'butterfly', 'bat', 'crab'
    X, A, B, C, D: tuple[int, float]  # Points with (index, price)
    direction: str  # 'bullish' or 'bearish'
    XA_retracement, AB_retracement, BC_retracement, CD_retracement: float
    prz: tuple[float, float]  # Potential Reversal Zone
    confidence: float

def detect_gartley(high, low, close, tolerance=0.02) -> list[HarmonicPattern]
def detect_butterfly(high, low, close, tolerance=0.02) -> list[HarmonicPattern]
def detect_bat(high, low, close, tolerance=0.02) -> list[HarmonicPattern]
def detect_crab(high, low, close, tolerance=0.02) -> list[HarmonicPattern]
def detect_harmonic_patterns(high, low, close, patterns=None, tolerance=0.02) -> list[HarmonicPattern]
Use standard Fibonacci ratios for each pattern type.

Utilities (utils.py)
def fibonacci_retracement(start: float, end: float) -> dict[str, float]
def fibonacci_extension(start: float, end: float, retracement: float) -> dict[str, float]
def fit_trendline(points: list[tuple[int, float]]) -> tuple[float, float]
def calculate_pattern_confidence(pattern_data: dict) -> float
2. Visualization Module (src/numta/viz/)
src/numta/viz/
├── __init__.py
├── lwcharts_adapter.py      # Integration with deepentropy/lwcharts
└── pattern_markers.py       # Pattern annotation helpers
def plot_chart(df, indicators=None, volume=True, height=400, width=800)
def plot_pattern(df, patterns, show_annotations=True, show_trendlines=True)
def plot_harmonic(df, patterns, show_fibonacci=True, show_prz=True)
3. Pandas Accessor Extension

Add to TAAccessor class in src/numta/pandas_ext.py:
def find_patterns(self, pattern_type='all', order=5, tolerance=0.03) -> list
def find_harmonic_patterns(self, patterns=None, tolerance=0.02) -> list[HarmonicPattern]
def plot(self, indicators=None, patterns=None, volume=True, height=400, width=800)
4. Update pyproject.toml
[project.optional-dependencies]
viz = ["pandas>=1.3.0", "lwcharts>=0.1.0"]
patterns = ["pandas>=1.3.0"]
full = ["pandas>=1.3.0", "lwcharts>=0.1.0", "numba>=0.56.0"]
5. Update src/numta/__init__.py

Export all pattern detection functions and dataclasses.

6. Tests

Create tests/test_patterns.py and tests/test_viz.py with comprehensive coverage.

7. Documentation

Update README.md with Pattern Recognition section including examples.

Acceptance Criteria

All chart patterns detected correctly with confidence scores

Harmonic patterns use correct Fibonacci ratios

Visualization works in Jupyter notebooks via lwcharts

Pandas .ta.find_patterns() and .ta.plot() methods work

Tests pass for all pattern types

Graceful degradation when lwcharts not installed

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.

Co-authored-by: deepentropy <8287111+deepentropy@users.noreply.github.com>

chatgpt-codex-connector · 2025-12-01T09:39:53Z

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.
To continue using code reviews, you can upgrade your account or add credits to your account and enable them for code reviews in your settings.

Copilot

Pull request overview

This PR implements Phase 2 pattern recognition and visualization capabilities for the numta library, adding comprehensive chart pattern detection with confidence scoring and optional visualization support via the lwcharts library.

Summary: The PR successfully adds pattern detection for classic chart patterns (Head & Shoulders, Double/Triple Top/Bottom, Triangles, Wedges, Flags, VCP) and harmonic patterns (Gartley, Butterfly, Bat, Crab) with Fibonacci ratio validation. It also provides visualization integration and extends the pandas accessor with .ta.find_patterns(), .ta.find_harmonic_patterns(), and .ta.plot() methods.

Key Changes:

New src/numta/patterns/ package with swing detection, pattern utilities, chart patterns, and harmonic patterns
New src/numta/viz/ package with lwcharts adapter and pattern marker utilities
Extended pandas accessor with pattern detection and plotting methods
Added optional dependencies for visualization and pattern features

Reviewed changes

Copilot reviewed 14 out of 14 changed files in this pull request and generated 13 comments.

Show a summary per file

File	Description
`src/numta/patterns/swing.py`	Swing high/low detection with support for higher/lower highs/lows
`src/numta/patterns/utils.py`	Fibonacci calculations, trendline fitting, pattern confidence scoring
`src/numta/patterns/chart_patterns.py`	Detection for 7 classic chart patterns with dataclasses
`src/numta/patterns/harmonic_patterns.py`	Detection for 4 harmonic patterns with standard Fibonacci ratios
`src/numta/viz/lwcharts_adapter.py`	Integration with lwcharts library with graceful degradation
`src/numta/viz/pattern_markers.py`	Helper functions for creating pattern annotations
`src/numta/pandas_ext.py`	Added `.ta.find_patterns()`, `.ta.find_harmonic_patterns()`, `.ta.plot()` methods
`tests/test_patterns.py`	58 new tests for pattern detection functionality
`tests/test_viz.py`	Comprehensive tests for visualization and marker creation
`pyproject.toml`	Added `viz`, `patterns`, and `full` optional dependency groups
`README.md`	Documentation for pattern recognition and visualization features

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

You can also share your feedback on Copilot code review for a chance to win a $100 gift card. Take the survey.

Copilot · 2025-12-01T09:44:01Z

tests/test_patterns.py

+        assert swing_lows[2] == True  # Trough at 8
+        assert swing_lows[6] == True  # Trough at 7


Avoid using explicit comparison with boolean literals. Instead of == True, use the value directly.

Replace:

assert swing_lows[2] == True with assert swing_lows[2]

assert swing_lows[6] == True with assert swing_lows[6]

Copilot · 2025-12-01T09:44:02Z

tests/test_patterns.py

+        assert price_within_tolerance(100.0, 102.0, 0.03) == True
+        assert price_within_tolerance(100.0, 110.0, 0.03) == False
+        assert price_within_tolerance(100.0, 100.0, 0.01) == True


Avoid using explicit comparison with boolean literals. Replace == True with the value directly.

Replace:

price_within_tolerance(100.0, 102.0, 0.03) == True with price_within_tolerance(100.0, 102.0, 0.03)

price_within_tolerance(100.0, 100.0, 0.01) == True with price_within_tolerance(100.0, 100.0, 0.01)

Suggested change

assert price_within_tolerance(100.0, 102.0, 0.03) == True

assert price_within_tolerance(100.0, 110.0, 0.03) == False

assert price_within_tolerance(100.0, 100.0, 0.01) == True

assert price_within_tolerance(100.0, 102.0, 0.03)

assert not price_within_tolerance(100.0, 110.0, 0.03)

assert price_within_tolerance(100.0, 100.0, 0.01)

Copilot · 2025-12-01T09:44:02Z

tests/test_viz.py

+        assert line['color'] == 'blue'
+        assert line['width'] == 2
+        assert line['style'] == 'dashed'
+        assert line['extend'] == True


Avoid using explicit comparison with boolean literals. Replace == True with the value directly.

Replace:

assert line['extend'] == True with assert line['extend']

Suggested change

assert line['extend'] == True

assert line['extend']

Copilot · 2025-12-01T09:44:02Z

tests/test_patterns.py

+        assert swing_highs[2] == True  # Peak at 15
+        assert swing_highs[6] == True  # Peak at 16
+        assert swing_highs[0] == False


Avoid using explicit comparison with boolean literals. Instead of == True, use the value directly. Instead of == False, use not value.

Replace:

assert swing_highs[2] == True with assert swing_highs[2]

assert swing_highs[0] == False with assert not swing_highs[0]

Suggested change

assert swing_highs[2] == True # Peak at 15

assert swing_highs[6] == True # Peak at 16

assert swing_highs[0] == False

assert swing_highs[2] # Peak at 15

assert swing_highs[6] # Peak at 16

assert not swing_highs[0]

Copilot · 2025-12-01T09:44:03Z

src/numta/viz/lwcharts_adapter.py

+
+        # Show PRZ zone if requested
+        if show_prz:
+            prz_lower, prz_upper = pattern.prz


Variable prz_lower is not used.

Suggested change

prz_lower, prz_upper = pattern.prz

Copilot · 2025-12-01T09:44:04Z

src/numta/patterns/chart_patterns.py

+"""
+
+from dataclasses import dataclass
+from typing import List, Optional, Tuple


Import of 'Optional' is not used.

Suggested change

from typing import List, Optional, Tuple

from typing import List, Tuple

Copilot · 2025-12-01T09:44:05Z

src/numta/viz/lwcharts_adapter.py

+If lwcharts is not installed, graceful degradation is provided.
+"""
+
+from typing import Any, Dict, List, Optional, Union


Import of 'Union' is not used.

Suggested change

from typing import Any, Dict, List, Optional, Union

from typing import Any, Dict, List, Optional

Copilot · 2025-12-01T09:44:05Z

src/numta/viz/pattern_markers.py

+and markers that can be used with various charting libraries.
+"""
+
+from typing import Any, Dict, List, Optional, Tuple, Union


Import of 'Union' is not used.
Import of 'Tuple' is not used.

Suggested change

from typing import Any, Dict, List, Optional, Tuple, Union

from typing import Any, Dict, List, Optional

Copilot · 2025-12-01T09:44:06Z

src/numta/viz/pattern_markers.py

+"""
+
+from typing import Any, Dict, List, Optional, Tuple, Union
+import numpy as np


Import of 'np' is not used.

Suggested change

import numpy as np

Copilot · 2025-12-01T09:44:07Z

tests/test_patterns.py

+from numta.patterns import (
+    # Swing detection
+    find_swing_highs,
+    find_swing_lows,
+    find_swing_points,
+    get_swing_high_indices,
+    get_swing_low_indices,
+    # Utilities
+    fibonacci_retracement,
+    fibonacci_extension,
+    fit_trendline,
+    price_within_tolerance,
+    calculate_pattern_confidence,
+    # Chart patterns
+    detect_head_shoulders,
+    detect_inverse_head_shoulders,
+    detect_double_top,
+    detect_double_bottom,
+    detect_triple_top,
+    detect_triple_bottom,
+    detect_triangle,
+    detect_wedge,
+    detect_flag,
+    detect_vcp,
+    # Harmonic patterns
+    detect_gartley,
+    detect_butterfly,
+    detect_bat,
+    detect_crab,
+    detect_harmonic_patterns,
+    # Dataclasses
+    HeadShouldersPattern,
+    DoublePattern,
+    TriplePattern,
+    TrianglePattern,
+    WedgePattern,
+    FlagPattern,
+    VCPPattern,
+    HarmonicPattern,
+)


Import of 'get_swing_low_indices' is not used.

Initial plan

b5523ca

Copilot AI assigned Copilot and deepentropy Dec 1, 2025

Copilot started work on behalf of deepentropy December 1, 2025 09:11 View session

Copilot AI and others added 2 commits December 1, 2025 09:27

Add pattern recognition and visualization modules (Phase 2)

79b904e

Co-authored-by: deepentropy <8287111+deepentropy@users.noreply.github.com>

Address code review feedback for patterns and viz modules

1b039d7

Co-authored-by: deepentropy <8287111+deepentropy@users.noreply.github.com>

Copilot AI changed the title ~~[WIP] Add chart pattern detection and visualization support~~ Add Pattern Recognition & Visualization (Phase 2) Dec 1, 2025

Copilot AI requested a review from deepentropy December 1, 2025 09:35

Copilot finished work on behalf of deepentropy December 1, 2025 09:35

deepentropy marked this pull request as ready for review December 1, 2025 09:39

Copilot AI review requested due to automatic review settings December 1, 2025 09:39

deepentropy merged commit 1ddc032 into main Dec 1, 2025
14 checks passed

deepentropy deleted the copilot/add-pattern-recognition-visualization branch December 1, 2025 09:39

Copilot started reviewing on behalf of deepentropy December 1, 2025 09:40 View session

Copilot finished reviewing on behalf of deepentropy December 1, 2025 09:41

Copilot AI reviewed Dec 1, 2025

View reviewed changes

		assert swing_lows[2] == True # Trough at 8
		assert swing_lows[6] == True # Trough at 7

	from typing import List, Optional, Tuple
	from typing import List, Tuple

	from typing import Any, Dict, List, Optional, Union
	from typing import Any, Dict, List, Optional

Add Pattern Recognition & Visualization (Phase 2) #21

Add Pattern Recognition & Visualization (Phase 2) #21

Uh oh!

Conversation

Copilot AI commented Dec 1, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

New Packages

src/numta/patterns/

src/numta/viz/

Pandas Accessor

Optional Dependencies

Tests

Phase 2: Pattern Recognition & Visualization

Overview

Deliverables

1. Chart Pattern Detection Module (src/numta/patterns/)

Swing Detection (swing.py)

Chart Patterns (chart_patterns.py)

Harmonic Patterns (harmonic_patterns.py)

Utilities (utils.py)

2. Visualization Module (src/numta/viz/)

3. Pandas Accessor Extension

4. Update pyproject.toml

5. Update src/numta/__init__.py

6. Tests

7. Documentation

Acceptance Criteria

Phase 2: Pattern Recognition & Visualization

Overview

Deliverables

1. Chart Pattern Detection Module (src/numta/patterns/)

Swing Detection (swing.py)

Chart Patterns (chart_patterns.py)

Harmonic Patterns (harmonic_patterns.py)

Utilities (utils.py)

2. Visualization Module (src/numta/viz/)

3. Pandas Accessor Extension

4. Update pyproject.toml

5. Update src/numta/__init__.py

6. Tests

7. Documentation

Acceptance Criteria

Uh oh!

chatgpt-codex-connector bot commented Dec 1, 2025

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Dec 1, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 1, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 1, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 1, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 1, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 1, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 1, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 1, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 1, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI Dec 1, 2025

Choose a reason for hiding this comment

Copilot AI commented Dec 1, 2025 •

edited

Loading

`src/numta/patterns/`

`src/numta/viz/`

1. Chart Pattern Detection Module (`src/numta/patterns/`)

Swing Detection (`swing.py`)

Chart Patterns (`chart_patterns.py`)

Harmonic Patterns (`harmonic_patterns.py`)

Utilities (`utils.py`)

2. Visualization Module (`src/numta/viz/`)

4. Update `pyproject.toml`

5. Update `src/numta/init.py`

1. Chart Pattern Detection Module (`src/numta/patterns/`)

Swing Detection (`swing.py`)

Chart Patterns (`chart_patterns.py`)

Harmonic Patterns (`harmonic_patterns.py`)

Utilities (`utils.py`)

2. Visualization Module (`src/numta/viz/`)

4. Update `pyproject.toml`

5. Update `src/numta/init.py`