knitr::opts_chunk$set(cache = TRUE, echo = TRUE, eval = FALSE)

Module metadata

Examine the LCC2005 group of modules:

  1. What versions of each of the modules are you using?

  2. Identify the timestep used for each module? How do they differ?

  3. Identify the input and output data (object) dependencies for these modules. Which are shared among them?

  4. What external data dependencies must be downloaded?

Module events

Examine the SpaDES sample modules:

  1. Identify the types of events found in each module. Which event types are "shared" among these modules?

  2. Compare the scheduling of plot events in each of the modules:

    a. what is the default plot interval for each module?

    b. at what times in the simulation are plot events scheduled and executed for each module? (hint: what is the default timeunit for each)

    ```r
    ## caribouMovement
    sim <- scheduleEvent(sim, time(sim) + P(sim)$.plotInterval,
                         "caribouMovement", "plot", .last())
    ```
    
    ```r
    ## fireSpread
    sim <- scheduleEvent(sim, time(sim) + P(sim)$.plotInterval,
                         "fireSpread", "plot", .last())
    ```
    
  3. Identify the input and output data (object) dependencies for these modules. Which are shared among them?

  4. What external data dependencies must be downloaded?

Debugging

Reminder: see the debugging info at the wiki.

Let's cause an error in an existing module:

  1. Create a copy of the gameOfLife module, called gameOfLifeError:

    ```r module.path <- file.path(dirname(tempdir()), "modules")

    hint: you'll need to download this module first

    downloadModule("gameOfLife", module.path)

    make a copy and open for editing

    copyModule("gameOfLife", "gameOfLifeError", module.path) openModules("gameOfLifeError", module.path) ```

  2. First, find and replace all instances of "gameOfLife" with "gameOfLifeError".

    Then, edit line 100 to read:

    r sim$world[r*f <= 2] <- FALSE

    and edit line 109 to read:

    r sim$world[!r*f == 3] <- TRUE

  3. Run the working module normally:

    ```r setPaths(modulePath = module.path)

    library(igraph) library(raster)

    X <- 10 Y <- 10 TYPE <- "blinker" ## see Rmd for other types

    modules <- list("gameOfLife") parameters <- list( gameOfLife = list(X = X, Y = Y, initialType = TYPE) ) times <- list(start = 1, end = 20)

    clearPlot() dev()

    mySim_OK <- simInit(times = times, params = parameters, modules = modules) mySim_OK1 <- spades(Copy(mySim_OK)) ```

  4. Run the broken version and confirm that you get an error:

    ```r X <- 10 Y <- 10 TYPE <- "blinker" ## see Rmd for other types

    modules <- list("gameOfLifeError") parameters <- list( gameOfLifeError = list(X = X, Y = Y, initialType = TYPE) ) times <- list(start = 1, end = 20)

    clearPlot() dev()

    mySim_ERR <- simInit(times = times, params = parameters, modules = modules) mySim_ERR1 <- spades(Copy(mySim_ERR)) ```

  5. Without knowing where the errors occur, how would you start debugging the error?

Create a new module

Create a simple module that does one of the following:

  1. generates a random map (hint: see ?gaussMap) and updates the values in the raster cells;

  2. model population dynamics as subpopulations within each raster cell.



PredictiveEcology/SpaDES.Workshops documentation built on Jan. 30, 2021, 6:52 p.m.