Outline

1 Introduction to ecological modelling with SpaDES

This workshop assumes good familiarity with R as well as several of its contributed packages.

2 Dates and Times:

Sept 14-16, 2016 google calendar link
8:30am - 4:30pm (PDT) each day
Cost: Free
Where: Pacific Forestry Centre, Victoria, BC
WebEx: Limited trial. Contact Eliot.McIntire@canada.ca if you would like to participate remotely.
Rooms:
- Wed: Dilbert Hall
- Thurs: Sage
- Fri: Sage

3 Workshop content

(raw version of notes available on github)

3.0.1 Before the course (link)

• Set up your laptop
• Set your goals for course
• all course material (incomplete until course begins), zipped

3.1 SpaDES in action (Eliot and Alex) (link)

• The demo modules in the SpaDES package
• LCC2005 model (“Land Cover Classification 2005”)
• Vegetation simulation (SpaDES-Landis)
• Agent based models – wolf IBM
• A shiny app on shinyapps.io (e.g., Proof of concept)

3.2 Thinking the SpaDES way (Eliot) (link)

• Events

• Modules

• Types of SpaDES modules:

  • events (e.g., Fire, Vegetation Change)
  • data preparation (e.g., climate data downloading)
  • individual-based modules (e.g., caribou, wolves, mountain pine beetle)
  • parents and children

• Data

3.3 Getting technical (Alex) (link)

1. The parts

• The simList

  • Modules
  • Events within modules
  • data

• The spades call

2. Surface dive

• creating the simList (simInit())
• run model (spades())
• where to get help
• using pre-built modules (downloadModule)

3.4 Building SpaDES modules (Alex) (link)

1. anatomy of a SpaDES module

2. new module template: newModule

3. module metadata defineModule

4. scheduling events: scheduleEvent

5. time

6. visualizations: Plot

7. debugging (spades(sim, debug = TRUE))

8. finding SpaDES tools

9. summary statistics

10. module development checklist

3.5 Simulation experiments and replication (Eliot) (link)

1. using the experiment() function for replication, scenario creation, and parameter experiments

2. running parallel simulations on supercomputers and clusters

3. Pattern Oriented Modeling (POM() function) for estimating unknown parameters

3.6 Getting the most out of R for ecological models (Eliot and Alex) (link)

1. Spatial data (raster and sp packages)

2. Matrices

3. The data.table package

4. SpaDES functions for spreading, moving, neighbourhoods etc. section 2 of spades-package help file

5. The Rcpp package

3.7 Module integration (Alex) (link)

1. Building “models”, i.e., groups of modules (parents and children)

2. Using metadata

3. Visual tools: objectDiagram, moduleDiagram, eventDiagram

3.8 Sharing modules & models (Alex) (link)

1. SpaDES module repositories

2. Using GitHub.com

3. shiny apps and shinyapps.io (e.g., Proof of concept)

4. Data sources

3.9 Data to decisions (Eliot) (link)

1. Building a reproducible workflow
2. Caching

4 Resources:

• SpaDES wiki
• R documentation for SpaDES
• Development release of SpaDES
• Link to this Outline
• Contact info: Eliot.McIntire@canada.ca or Alex.Chubaty@canada.ca

5 Future offering

Dec 7-9, 2016 (tentative)
Depending on how the WebEx goes in September, this offering may be offered remotely.