easyshapey 0.0.5

package to draw boxes in 2d plots

easyshapey

A tool for manipulating 2D shapes in plots. Select data points, create arbitrary polygons, and visualize shapes on matplotlib axes.

Installation

pip install easyshapey

Or from source:

pip install git+https://github.com/caganze/easyshapey.git

Features

• Box: Rectangular shapes fitted to data trends
• RotatedBox: Auto-rotated boxes minimizing bounding area
• Oval: Elliptical shapes fitted to data
• Polygon: Arbitrary N-sided polygons (triangles to any N)

All shapes support:

• Data point selection (select())
• Point containment checking (contains())
• Rotation (rotate())
• Plotting on matplotlib axes (plot())

Quick Start

Box

import numpy as np
import pandas as pd
from easyshapey import Box

# Create data
x = np.random.random(100)
y = 2 * x + np.random.normal(0, 0.2, 100)
data = np.array([x, y])

# Fit box to data
box = Box()
box.data = data

# Select points inside
selected = box.select(data)

# Plot
box.plot()

Polygon

from easyshapey import Polygon

# Create triangle
tri = Polygon(vertices=[(0, 0), (1, 0), (0.5, 1)])
print(f"Sides: {tri.n_sides}, Area: {tri.area:.2f}")

# Create any N-sided polygon
hexagon = Polygon(vertices=[
    (1, 0), (0.5, 0.87), (-0.5, 0.87),
    (-1, 0), (-0.5, -0.87), (0.5, -0.87)
])

# Check point containment
inside = tri.contains([(0.5, 0.3)])  # [True]

# Rotate 45 degrees
tri.rotate(np.pi / 4)

Interactive Polygon Creation

from easyshapey import Polygon

# Click points on figure, press 'q' when done
poly = Polygon.from_clicks(min_points=3)

Polygon from Data

from easyshapey import Polygon
import numpy as np

data = np.random.rand(2, 50)  # 50 random points

# Bounding box
bbox = Polygon.from_data(data, method='bounding_box')

# Convex hull (requires scipy)
hull = Polygon.from_data(data, method='convex_hull')

Selector

from easyshapey import Box, Selector

# Combine multiple shapes
box1 = Box()
box1.data = data1

box2 = Box()
box2.data = data2

selector = Selector()
selector.shapes = [box1, box2]
selector.logic = 'and'  # or 'or'

# Select points in all shapes
result = selector.select()

API Reference

Polygon

Property	Description
n_sides	Number of sides
vertices	List of (x, y) tuples
center	Centroid (x, y)
area	Area (shoelace formula)
angle	First edge angle (radians)

Method	Description
contains(points)	Check if points inside
select(data)	Select data points inside
rotate(angle)	Rotate around center
plot(**kwargs)	Plot on matplotlib axes
from_clicks()	Create interactively
from_data()	Create from data points

Box

Property	Description
vertices	Corner vertices
center	Center (x, y)
area	Bounding area
coeffs	Fitted line [slope, intercept]

Examples

See the tutorial notebook for more examples.

---

Changelog

[0.0.4] - 2025-01-07

Added

• Polygon class: arbitrary N-sided 2D polygons
• Interactive creation via Polygon.from_clicks()
• Data fitting via Polygon.from_data()
• Comprehensive test suite

Changed

• Streamlined code following Abseil performance principles
• Consistent NumPy-style docstrings
• ~40% leaner implementation

[0.0.3] - 2025-01-07

Security

• Fixed deprecated pandas .ix[] with .loc[]
• Updated security policy

[0.0.2] - 2022-05-04

• Version bump and package updates

[0.0.1] - 2022-01-27

• Initial release

---

Security

Supported Versions

Version	Supported
0.0.4	✓
0.0.3	✓
< 0.0.3	✗

Reporting Vulnerabilities

Email: [e-mail]

Include: description, reproduction steps, impact, suggested fix.

Response time: 48 hours acknowledgment, 7 days initial assessment.

---

License

MIT License

Author

caganze - [e-mail]

Links

• PyPI
• GitHub