Points

SetPoints

Overview

The SetPoints class represents an ordered collection of points in n-dimensional Euclidean space. It provides various methods for generating, transforming, and visualizing point sets.

Class: SetPoints

Represent an ordered collection of points in n-dimensional space.

Attributes:

• n (int): The number of points in the collection.

• dim (int): The dimensionality of the points.

• points (numpy.ndarray): A 2-dimensional numpy array containing the point coordinates with shape (n, dim).

• pos (dict): A dictionary mapping point indices to their coordinates.

• centroid (numpy.ndarray): The centroid (mean position) of all points.

Constructor

__init__(points, seed=None) …

Base constructor for SetPoints.

Parameters:

• points (numpy.ndarray): A 2D numpy array of shape (n, dim) where n is the number of points and dim is the dimension of each point.

• seed (int, optional): A seed for the random number generator to ensure reproducibility. If None, a default random generator is used.

Raises

• TypeError: If points is not a numpy.ndarray.

Returns

A SetPoints object containing the points.

Example

import numpy as np
from proximitygraphs.points import SetPoints

# Create from numpy array
points_array = np.random.rand(100, 2)
my_points = SetPoints(points_array, seed=42)

—

This is a basic overview of the SetPoints class. For detailed information on methods and additional functionalities, please refer to the specific method documentation.

Submodules

• Uniform square
• uniform_over_sphere
• uniform_sphere
• normal_dist
• grid
• hexagonal
• triangular
• poissonprocess_square
• poissonprocess_circle
• poissonprocess_inhomogeneus
• cluster_square
• from_geopandas
• Transformation Methods
• Special Methods
• Draw

---

Usage Tips

• Always use seed for reproducibility.

• draw() is only available for 2D data.

• Transformations are immutable: they return new objects.

• Large datasets may be memory-intensive.

• from_geopandas() integrates with GeoPandas workflows.

---

Common Workflows

Generate and Transform Points

from proximitygraphs.points import SetPoints

points = SetPoints.uniform_square(n=100, dims=2, seed=42)

scaled = points.scaling(2.0)
rotated = scaled.rotation(45)
translated = rotated.traslation([1, 1])
noisy = translated.perturb(radius=0.1)

noisy.draw(figsize=(10, 10), v_color='blue', v_alpha=0.6)

Create Custom Point Patterns

from proximitygraphs.points import SetPoints

uniform_bg = SetPoints.uniform_square(n=200, dims=2, seed=1)
clusters = SetPoints.cluster_square(
    intensity=(3, 15),
    cluster={"name": "Thomas", "param": 0.05},
    limit=1,
    seed=2
)

combined = uniform_bg + clusters
combined.draw(figsize=(10, 10), v_color='red', v_size=15)

—