Fit the spread model

The module's one fireSenseUtils::runDEoptim() call, shared by the run event (every year) and crossValidate (one fold's years). After the fit, runDEoptim() re-scores the final population and, with profileReps/simulateMembers above 0, profiles the best member and simulates the best members' fires, on the same workers; makeFitDiagnostics() turns those into sim objects.

fitSpread(sim, covs, thresh, runName, diagnostics = TRUE)

Arguments

sim

a simList.

covs

the covariates as integers x 1000 (mod$covsX1000, or a subset of its years).

thresh

the objective's early-stop threshold, mod$thresh.

runName

character; labels the run and its cache entry.

diagnostics

logical; FALSE skips the profile and the simulations.

Value

the runDEoptim() result.