fireSense_SpreadPredict.Rmd

---
title: "fireSense_SpreadPredict"
author:
  - "Jean Marchal (<jean.d.marchal@gmail.com>)"
  - "Alex M. Chubaty (<achubaty@for-cast.ca>)"
  - "Tati Micheletti (<tati.micheletti@gmail.com>)"
date: "`r format(Sys.time(), '%d %B %Y')`"
output:
  html_document: default
editor_options: 
  chunk_output_type: console
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE, results = "hold")
```

# Overview

Predicts a surface (raster) of fire spread probabilities using a model previously fit with `fireSense_SpreadFit`.

# Usage

```{r module-usage-init}
library(magrittr)
library(raster)
library(SpaDES.core)
library(SpaDES.tools)

set.seed(123)

# Define simulation parameters
times <- list(start = 1, end = 1, timeunit = "year")
modules <- list("fireSense_SpreadPredict")
paths <- list(
  modulePath = ".."
)

# Create a random map of weather
nx <- ny <- 100L
weather <- raster(nrows = ny, ncols = nx, xmn = -nx/2, xmx = nx/2, ymn = -ny/2, ymx = ny/2) %>%
  gaussMap(scale = 300, var = 1, speedup = 1, inMemory = TRUE)

# Create a typical output of fireSense_SpreadFit
fireSense_SpreadFitted <- list(
  formula = ~ weather2 - 1,
  coef = setNames(c(0.3, 3, 0.1, 1.5, 3),
                  c("A", "B", "D", "G", "weather"))
  ## A, B, D, G are parameters of the 5-parameters logistic function
)
class(fireSense_SpreadFitted) <- "fireSense_SpreadFit"

parameters <- list(
  fireSense_SpreadPredict = list(
    mapping = list(weather2 = "weather") # One can use mapping to map variables
                                         # in the formula of the fitted object
                                         # to those in data. Here weather2
                                         # (formula) is mapped to weather (data).
  )
)

# Objects to pass from the global environment to the simList environment
objects <- c(
  "weather" = weather,
  "fireSense_SpreadFitted" = fireSense_SpreadFitted
)

# Create the simList
sim <- simInit(
  times = times,
  modules = modules,
  paths = paths,
  params = parameters,
  objects = objects
)
```

Run the model: <!-- TODO: this currently FAILS due to incorrect object class -->

```{r module-usage-run, eval = FALSE}
sim <- spades(sim)

Plot(weather, sim$fireSense_SpreadPredicted)
```

# Parameters

Provide a summary of user-visible parameters.

```{r moduleParams, echo = FALSE}
df_params <- moduleParams("fireSense_SpreadPredict", "..")
knitr::kable(df_params)
```

# Events

Events are scheduled as follows:

- Module initiation;
- Make predictions.

## Saving

There is currently nothing saved, but this may change in the future. <!-- TODO -->

# Data dependencies

## Input data

```{r moduleInputs, echo = FALSE}
df_inputs <- moduleInputs("fireSense_SpreadPredict", "..")
knitr::kable(df_inputs)
```

## Output data

```{r moduleOutputs, echo = FALSE}
df_outputs <- moduleOutputs("fireSense_SpreadPredict", "..")
knitr::kable(df_outputs)
```

# Links to other modules

Predictions made with this module can be used with the fire spread component of landscape fire models (e.g., `fireSense`). <!-- TODO -->