pyehsa 0.1.1

A Python library for Emerging Hot Spot Analysis (EHSA) of spatio-temporal data

PyEhsa: Emerging Hot Spot Analysis in Python

PyEhsa is a Python library for Emerging Hot Spot Analysis (EHSA) of spatio-temporal data, providing functionality similar to R's sfdep package. It enables researchers and analysts to identify and classify spatial-temporal patterns, emerging hotspots, and coldspots in geographic data.

Key Features

• Emerging Hot Spot Analysis: Identify emerging patterns in spatio-temporal data
• Spatial Statistics: Calculate Getis-Ord Gi* statistics for hotspot detection
• Mann-Kendall Trend Analysis: Detect trends in time series data
• Multiple Classifications: Comprehensive hotspot pattern classification
• Interactive Visualization: HTML tool with time series analysis and mapping
• Flexible Data Input: Support for pandas DataFrames and GeoPandas GeoDataFrames

Installation

pip install pyehsa

Requirements

• Python 3.9+
• pandas >= 1.5.0
• geopandas >= 0.13.0
• numpy >= 1.21.0
• scipy >= 1.9.0
• libpysal >= 4.6.0

Quick Start Guide

Step 1: Import Libraries

import pandas as pd
import geopandas as gpd
from datetime import datetime, timedelta
from pyehsa import EmergingHotspotAnalysis

Step 2: Prepare Your Data

Your dataset should have:

• region_id: Spatial identifier (e.g., 'location_001', 'geohash_abc')
• time_period: Temporal column (datetime format)
• value: Numeric variable to analyze
• geometry: Spatial geometry (Point or Polygon from shapely)

# Example: Create or load your GeoDataFrame
data = pd.DataFrame({
    'region_id': ['area_1', 'area_1', 'area_2', 'area_2'],
    'time_period': [datetime(2024, 1, 1), datetime(2024, 2, 1), 
                    datetime(2024, 1, 1), datetime(2024, 2, 1)],
    'value': [10.5, 15.2, 8.3, 12.1],
    'geometry': [polygon_1, polygon_1, polygon_2, polygon_2]  # Your geometries
})

gdf = gpd.GeoDataFrame(data, geometry='geometry', crs='EPSG:4326')

Step 3: Run Emerging Hot Spot Analysis

results = EmergingHotspotAnalysis.emerging_hotspot_analysis(
    gdf,
    region_id_field='region_id',
    time_period_field='time_period',
    value='value',
    k=1,      # Temporal lags to include
    nsim=99   # Monte Carlo simulations for significance testing
)

Parameter Details:

• region_id_field (str): Name of the column containing spatial region identifiers. This column uniquely identifies each geographic area in your analysis (e.g., 'geohash_6', 'zip_code', 'grid_id').

• time_period_field (str): Name of the column containing temporal information. Must be in datetime format. This defines the time periods for the spatio-temporal analysis (e.g., 'month', 'date', 'time_period').

• value (str): Name of the column containing the numeric values to analyze. This is the variable you want to detect hotspots for (e.g., 'crime_count', 'sales', 'temperature').

• k (int, default=1): Number of temporal lags to include in the analysis. This defines how many previous time periods should be considered as temporal neighbors.

  • k=1: Only includes the immediate previous time period (most common)
  • k=2: Includes the two previous time periods
  • k=3: Includes the three previous time periods
  • Higher k values capture longer temporal dependencies but require more time periods in your data

• nsim (int, default=99): Number of Monte Carlo simulations for statistical significance testing. Higher values provide more robust p-values but increase computation time.

  • nsim=99: Fast, suitable for exploratory analysis
  • nsim=199: Balanced (matches R's default)
  • nsim=999: High precision for final analyses
  • Minimum recommended: 99

Step 4: View Results

# See classification distribution
print(results['classification'].value_counts())

# View detailed results
results.head()

Step 5: Create Interactive Visualization

from pyehsa import EhsaPlotting

# Merge geometry with results for visualization
locations = gdf[['region_id', 'geometry']].drop_duplicates()
viz_data = results.merge(locations, left_on=results.columns[0], right_on='region_id')

# Create interactive map
ehsa_map = EhsaPlotting.plot_ehsa_map_interactive(
    df=viz_data,
    region_id_field='region_id',
    title="Emerging Hotspots Analysis"
)
ehsa_map.show()

Classification Types

PyEhsa identifies 17 different spatio-temporal patterns:

Hotspot Patterns:

• new hotspot: Recently emerged statistically significant hotspot (significant only in final time period)
• consecutive hotspot: Hotspot significant for multiple consecutive recent periods with <90% overall significance
• intensifying hotspot: Hotspot significant in ≥90% of time periods with statistically significant increasing trend
• persistent hotspot: Hotspot significant in ≥90% of time periods with no significant trend
• diminishing hotspot: Hotspot significant in ≥90% of time periods with statistically significant decreasing trend
• sporadic hotspot: On-and-off hotspot with some significant periods but no clear pattern
• oscillating hotspot: Hotspot that alternates between significant and non-significant with majority being significant
• historical hotspot: Only significant in ≥90% of early time periods but not recent periods

Coldspot Patterns:

• new coldspot: Recently emerged statistically significant coldspot
• consecutive coldspot: Coldspot significant for multiple consecutive recent periods
• intensifying coldspot: Coldspot significant in ≥90% of periods with significant increasing intensity trend
• persistent coldspot: Coldspot significant in ≥90% of periods with no significant trend
• diminishing coldspot: Coldspot significant in ≥90% of periods with significant decreasing intensity trend
• sporadic coldspot: On-and-off coldspot with intermittent significant periods
• oscillating coldspot: Coldspot that alternates between significant and non-significant states
• historical coldspot: Only significant in early time periods but not recent

No Pattern:

• no pattern detected: Areas without statistically significant spatial-temporal patterns

Advanced Usage

Interactive HTML Visualization Tool

PyEhsa includes a powerful interactive HTML visualization tool for exploring your EHSA results:

# Step 1: Save your results to CSV
results.to_csv('ehsa_results.csv', index=False)

# Step 2: Launch the interactive visualization tool
from pyehsa import EhsaPlotting
EhsaPlotting.open_visualization_tool()

This will open an interactive HTML tool in your browser where you can:

1. Upload your CSV: Click "Choose File" and select ehsa_results.csv
2. Explore the map: Interactive visualization of all spatial patterns
3. Inspect time series: Click any region and select "Inspect Time Series" to see temporal trends
4. View statistics: Detailed Mann-Kendall statistics and classification details for each region

The visualization tool provides deep insights into your spatio-temporal patterns with interactive charts and detailed region analysis.

Complete Example

See examples/demo_sp.ipynb for a complete working example with:

• Synthetic grid data covering São Paulo
• Full EHSA workflow from data creation to visualization
• Interactive map generation

Contributing

We welcome contributions! Please:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

Citation

If you use PyEhsa in your research, please cite:

@software{pyehsa2025,
  author = {CloudWalk},
  title = {PyEhsa: Emerging Hot Spot Analysis in Python},
  year = {2025},
  url = {https://github.com/cloudwalk/PyEhsa},
  version = {0.1.0}
}

License

This project is licensed under the MIT License - see the LICENSE file for details.

Related Work

PyEhsa is inspired by and provides similar functionality to:

• R's sfdep package for spatial dependence analysis
• ArcGIS's Emerging Hot Spot Analysis tool
• PySAL ecosystem for spatial analysis

Development & Support

• Repository: github.com/cloudwalk/PyEhsa
• Issues: GitHub Issues
• PyPI: pypi.org/project/pyehsa

Acknowledgments

Developed by CloudWalk's data science team to provide open-source spatial analysis tools for the Python community.