Technical Workshops

Extending ArcGIS with R

Mark Janikas, Ph. D.

Outline

• Introduction via demonstration- Analyzing the future of the Wolverine in

Washington State

• Description of R- Why/When to use it?

• Integration options- How to use it?

• Conclusions and Future Directions

Last Year: Point Clustering (Partitioning)

• Cluster a given a set of point locations:- Spatial Proximity- Attributes Values (Including Time)- Cancer Clusters in New Mexico - Partitioning Earthquakes in California

The Problem

• We want to examine what the potential effects of climate change will be on the distribution of animal species

• We have the known current locations of the distributions of the species

• We have a series of independent variables including- Vegetation type (as dummy variables)

- Elevation, slope, and aspect

- Distance from roads and cities

- Etc.

The climate data

• We have two climate change models - Hadley (from the UK)

- MIROC 3.2 (from Japan)

• Each model has two scenarios- The moderate, mid-level “A1B” carbon scenario

- The higher, more extreme “A2” carbon scenario

• There are three time periods- “e”: Early-century, or 2020-2024 averaged

- “m”: Mid-century, or 2050-2054 averaged

- “l”: Late-century, or 2095-2099 averaged

From Ron Nielson’s group at Oregon State University/ US Forest Service

Special Thanks to Kevin Johnston

The model

• Sample points created from the raster data

• Tools created to run R logistics regression

• Fit model

• Coefficients and diagnostics statistics

• Use coefficients to create a raster surface

Creating the raster surface

• Apply the logistics formula with coefficients

• Select for probability of .5 or greater

• Repeat for each model, for each scenario, and for time period

1 / (1 + exp( -1 * (9.595857 + (-1.28212 * tmp1991) + (-0.003687 * ppt1991) + (0.426121 * veg8_10) + (-0.560821 * veg7_10) + (-2.077026 * veg6_10) + (-2.941375 * veg2_10) + (-0.496024 * veg17_10) + (-1.740473 * veg16_10) + (0.557113 * veg12_10) + (-7.103907 * veg10_10) + (0.016223 * slope) + (-0.000674 * elevation) + (-0.000555 * aspect) + (-0.000062 * disthigh) + (0.000049 * distcity))))

Demo Title: Climate Change

Sampling

Logistics regression

Creating a surface

What is R? Why should I use it?

• R (The R Project for Statistical Computing) is an open-source data analysis package. (GNU S)- Widely Used

- Over 60 CRAN sites across 30+ countries

- Its Free- GNU GENERAL PUBLIC LICENSE

- Base is powerful- Statistics, Linear Algebra, Visualization , etc…

- Its extendible- 1800+ Contributed Extensions

- splancs, spatstat, spdep, rgdal, maptools, shapefiles

Integration with ArcGIS

• Two Integration Options With ArcGIS

• Both require Python

• Both have pros and cons

• ESRI UC Plenary 2008- predicting plant species in unknown areas

Integration: R Option

• Decouples R and Python

• Python - Retrieves and organize parameters from ArcGIS

- Convert Data (Interchange)- Shapefiles, netcdf, img etc….

- Spawns R given the *.r file with provided parameters

• R- Does the analysis

• Python- Post-Processing

- Projecting Data, Applying Symbology

Integration: R Option (Cont…)

ArcGISArcGIS R ScriptR ScriptPython ScriptPython Script

Integration: RPy Option

• R and Python closely coupled

• RPy and RPy2- Python Interface to the R Programming Language

• Python- Retrieves and organize parameters from ArcGIS

- RPy module is imported and R commands are executed within the Python script file

Integration: RPy Option (Cont…)

ArcGISArcGIS

Python ScriptPython Script

R ProcessingR Processing

Which One Should I Use?

• R Option vs RPy (RPy2) Options

• http://resources.arcgis.com/gallery/file/geoprocessing

• ArcGIS 9.3 vs. ArcGIS 10

Which One Should I Use? (Cont…)

• R Option- Integration Easy to Implement

- Attractive to R Programmers

- “Out of Proc”: Spawning R on every execute

- Use Copy Features- Shapefiles

- selection sets

- Projections and other environment variables

- You must use an R library for handling shapefiles- maptools, shapefiles

- Two files per script tool (*.py and *.r)

- Debugging can be difficult

Which One Should I Use? (Cont…)

• RPy / RPy2 Option- For more advanced users (Python and R knowledge)

- “In Process”- Will be MUCH faster after the first call

- Honors selection sets

- A robust choice of database formats

- Will honor environment settings (GP Functions)

- Only a single file associated with your script tool

- RPy (First Generation)- Existing Bug with GP Python Framework

- RPy2 (“Second” Generation 2.0.x)- Interaction with NumPy arrays incomplete

- RPy2 (2.1.x)- No Windows Binaries

•

Conclusions

• R - contains “cutting edge” data analysis techniques from

a wide body of academic and applied fields

- extendible

- Open-source

• Can be integrated with ArcGIS using Python- R versus RPy (RPy2)

- Pros and Cons

Future Directions

• RPy2

• Web Portal: RTools- spdep, spatstat and splancs

- SAR/CAR Regression, Point Pattern Analysis

• ArcNews Article

• MATLAB, SPSS, PySAL

• Calling Python from R- Leveraging geoprocessing within the R environment

- RSPython: http://www.omegahat.org/RSPython/

Links

• R- http://www.r-project.org/index.html

• Rpy and RPy2- http://rpy.sourceforge.net/

• Python - http://www.python.org/

• NumPy - http://www.numpy.org/

Related Sessions

• Spatial Pattern Analysis

• Regression Analysis for Spatial Data with ArcGIS

• Geostatistical Analyst

• ArcGIS Spatial Analyst – Statistical Modeling

• Agent-Based Modeling in ArcGIS

• Python Essentials in ArcGIS