Run an epicrop Disease Model Across Emergence Dates and Locations

A unified interface to execute any epicrop disease model (bacterial_blight(), brown_spot(), leaf_blast(), etc.) over multiple emergence dates and/or locations. Handles cross-year and within-year weather windows automatically without global weather binding.

Usage

run_epicrop_model(
  model_fun,
  emergence_dates,
  wth_list,
  window_days = 120L,
  output = "audpc",
  ...
)

Arguments

model_fun

Function. An epicrop model function such as: bacterial_blight(), brown_spot(), leaf_blast(), sheath_blight(), tungro(), modified_kim_leaf_blast(), modified_kim_sheath_blight(), leaf_rust(), or s_tritici_blotch().

emergence_dates

Vector of emergence dates (character "YYYY-MM-DD", Date, or data.table::IDate; must be of class epicrop.emerge from build_epicrop_emergence()).

wth_list

Named list of weather data.table::data.tables, one per year.

Names must be year strings (e.g., "2001", "2002")
Each table must have a YYYYMMDD column (Date, IDate, or ISO string)
Typically created by fetch_epicrop_weather_list()
Can also be an epicrop.wth.bundle (multiple locations)

window_days

Integer. Days after emergence to simulate (default 120L). The model will run for exactly window_days rows when sufficient weather is available. If emergence_dates is at year-end, spans into next year.

output

Character. Output format:

"audpc" (default): Compact summary with emergence date and AUDPC
"full": All daily model columns plus emergence and AUDPC

...

Additional arguments forwarded to model_fun (e.g., RcOpt, simple_wetness, custom modifier tables). Overrides model defaults.

Value

A data.table::data.table:

If output = "audpc":
- Columns: emergence (IDate), AUDPC (numeric)
- One row per emergence date
- AUDPC = NA_real_ if weather data are insufficient
If output = "full":
- All columns from the disease model (sites, latent, infectious, etc.)
- Plus emergence (IDate) and AUDPC (numeric)
- Multiple rows per emergence date (one per simulated day)
- Class set to epicrop.sim for S3 methods
If wth_list is a bundle (multiple locations):
- Additional location column identifying source bundle member

Workflow

This function bridges fetch_epicrop_weather_list() (weather prep) and build_epicrop_emergence() (emergence date generation) with disease models. For each emergence date, it:

Slices weather data spanning emergence to emergence + window_days - 1
Handles year boundaries (e.g., Dec 15 – Jan 15)
Executes the model with the sliced weather
Returns either a compact AUDPC summary or full daily progression

Supports weather bundles (multiple locations) by processing each location independently and combining results with a location column.

Cross-Year Windows

When an emergence date + window_days spans two calendar years (e.g., Dec 15 – Feb 15), the function automatically:

Fetches weather from both years
Stitches them in chronological order
Returns a contiguous window

Missing data (no weather for a required year) produces a warning and AUDPC = NA_real_.

Bundle Processing

If wth_list is an epicrop.wth.bundle (from fetch_epicrop_weather_list() with multiple locations or location-date pairs):

Each bundled weather object is processed independently
Results are combined with a location column
Location labels extracted from bundle names (text before "__")

Author

Adam H. Sparks, adamhsparks@gmail.com

Examples

# --- Example 1: Single location, multiple emergence dates ---

# Build emergence dates (every 15 days, June–July, years 2001–2003)
years <- 2001:2003
month_days <- c("-06-01", "-06-15", "-07-01")
emergence <- build_epicrop_emergence(years, month_days)

# Fetch weather (one year per element)
lonlat <- c(121.255669, 14.16742)
wth_list <- fetch_epicrop_weather_list(
  lonlat = lonlat,
  start_date = month_days,
  duration = 120L,
  years = years
)

# Run leaf blast model across all emergence dates
results <- run_epicrop_model(
  model_fun = leaf_blast,
  emergence_dates = emergence,
  wth_list = wth_list,
  window_days = 120L,
  output = "audpc"
)
head(results)
#> Key: <emergence, AUDPC>
#>           emergence    AUDPC
#>    <epicrop.emerge>    <num>
#> 1:       2001-06-01 38.19575
#> 2:       2001-06-15 36.86150
#> 3:       2001-07-01 43.86372
#> 4:       2002-06-01 40.34144
#> 5:       2002-06-15 38.03268
#> 6:       2002-07-01 37.42458

# --- Example 2: Full progression output for detailed analysis ---

full_results <- run_epicrop_model(
  model_fun = leaf_blast,
  emergence_dates = emergence[1:5],  # First 5 dates
  wth_list = wth_list,
  window_days = 120L,
  output = "full"
)

# Plot disease progression for first emergence date
first_date <- emergence[1]
sub <- full_results[emergence == first_date]
plot(sub$simday, sub$intensity, type = "l",
     main = paste("Leaf Blast:", first_date),
     xlab = "Days After Emergence", ylab = "Disease Intensity")


# --- Example 3: Sensitivity analysis (modify RcOpt) ---

# Susceptible cultivar
susceptible <- run_epicrop_model(
  model_fun = leaf_blast,
  emergence_dates = emergence,
  wth_list = wth_list,
  window_days = 120L,
  output = "audpc",
  RcOpt = 1.14  # Default
)
#> Error in seir(wth = wth, emergence = emergence, RcA = RcA %||% params$RcA,     RcT = RcT %||% params$RcT, RRG = RRG %||% params$RRG, RRS = RRS %||%         params$RRS, onset = 15L, duration = 120L, rhlim = rhlim %||%         90L, rainlim = rainlim %||% 5L, H0 = 600L, I0 = 1L, RcOpt = 1.14,     p = 5L, i = 20L, a = 1L, Sx = 30000L, simple_wetness = TRUE,     age_driver = "day", ...): formal argument "RcOpt" matched by multiple actual arguments

# Resistant cultivar (lower infection rate)
resistant <- run_epicrop_model(
  model_fun = leaf_blast,
  emergence_dates = emergence,
  wth_list = wth_list,
  window_days = 120L,
  output = "audpc",
  RcOpt = 0.8  # Lower
)
#> Error in seir(wth = wth, emergence = emergence, RcA = RcA %||% params$RcA,     RcT = RcT %||% params$RcT, RRG = RRG %||% params$RRG, RRS = RRS %||%         params$RRS, onset = 15L, duration = 120L, rhlim = rhlim %||%         90L, rainlim = rainlim %||% 5L, H0 = 600L, I0 = 1L, RcOpt = 1.14,     p = 5L, i = 20L, a = 1L, Sx = 30000L, simple_wetness = TRUE,     age_driver = "day", ...): formal argument "RcOpt" matched by multiple actual arguments

comparison <- data.table(
  emergence = susceptible$emergence,
  susceptible_AUDPC = susceptible$AUDPC,
  resistant_AUDPC = resistant$AUDPC
)
#> Error: object 'susceptible' not found
comparison[, reduction := 1 - (resistant_AUDPC / susceptible_AUDPC)]
#> Error: object 'comparison' not found
comparison
#> Error: object 'comparison' not found

# --- Example 4: Bundle processing (multiple locations) ---

locations <- rbind(
  "IRRI" = c(121.255669, 14.16742),
  "Laguna" = c(121.2, 14.1)
)

wth_bundle <- fetch_epicrop_weather_list(
  lonlat = locations,
  start_date = month_days,
  duration = 120L,
  years = years,
  mode = "cross"  # All location × date combinations
)

bundle_results <- run_epicrop_model(
  model_fun = leaf_blast,
  emergence_dates = emergence,
  wth_list = wth_bundle,
  window_days = 120L,
  output = "audpc"
)

# Results now have a 'location' column
bundle_results[, .N, by = location]
#>    location     N
#>      <char> <int>
#> 1:     IRRI     9
#> 2:   Laguna     9