## Warning: `as_tibble.matrix()` requires a matrix with column names or a `.name_repair` argument. Using compatibility `.name_repair`.
## This warning is displayed once per session.

Summer

Summer DFO model includes SST, bathymetry, and calanus.

summer.model <- dismo::maxent(summerEnv, summer_train)

saveRDS(summer.model,file = "summer_mod.RDS")
# plot showing importance of each variable
plot(summer.model)

# response curves
response(summer.model)

# predict to entire dataset
summer_pred_vals <- dismo::predict(summer.model, summerEnv, ext = model.extent)

#plot predictions
plot(summer_pred_vals, main="Predicted Suitability")
maps::map('worldHires', fill=FALSE, add=TRUE)
points(summer_train$long, summer_train$lat, pch=16, cex=1, col = rgb(red = 0, green = 0, blue = 1, alpha = 0.1))

Winter

Winter DFO model includes SST.

winter.model <- maxent(winterEnv, winter_train)
# plot showing importance of each variable
plot(winter.model)

# response curves
response(winter.model)

# predict to entire dataset
winter_pred_vals <- dismo::predict(winter.model, winterEnv)

#plot predictions
plot(winter_pred_vals, main="Predicted Suitability")
maps::map('worldHires', fill=FALSE, add=TRUE)
points(winter_train$long, winter_train$lat, pch=16, cex=1, col = rgb(red = 0, green = 0, blue = 1, alpha = 0.1))

Spring

Spring DFO model includes SST, bathymetry, and calanus.

spring.model <- maxent(springEnv, spring_train)
# plot showing importance of each variable
plot(spring.model)

# response curves
response(spring.model)

# predict to entire dataset
spring_pred_vals <- dismo::predict(spring.model, springEnv)

#plot predictions
plot(spring_pred_vals, main="Predicted Suitability")
maps::map('worldHires', fill=FALSE, add=TRUE)
points(spring_train$long, spring_train$lat, pch=16, cex=1, col = rgb(red = 0, green = 0, blue = 1, alpha = 0.1))

Fall

Summer DFO model includes SST and calanus.

fall.model <- maxent(fallEnv, fall_train)
# plot showing importance of each variable
plot(fall.model)

# response curves
response(fall.model)

# predict to entire dataset
fall_pred_vals <- dismo::predict(fall.model, fallEnv)

#plot predictions
plot(fall_pred_vals, main="Predicted Suitability")
maps::map('worldHires', fill=FALSE, add=TRUE)
points(fall_train$long, fall_train$lat, pch=16, cex=1, col = rgb(red = 0, green = 0, blue = 1, alpha = 0.1))

#save(summer.model,spring.model,fall.model,winter.model,summerEnv,springEnv,fallEnv,winterEnv,
     #summer_pred_vals,spring_pred_vals,fall_pred_vals,winter_pred_vals,
     #summer_test, summer_train, spring_test, spring_train, fall_test, fall_train, 
    # winter_test, winter_train, file = "maxentmods.RData")
## [1] 0.6988654 0.6988654 0.6988654 0.6988654 0.6988654 0.7696608
## class          : ModelEvaluation 
## n presences    : 790 
## n absences     : 151 
## AUC            : 0.8383435 
## cor            : 0.7030698 
## max TPR+TNR at : 0.6987654
##                kappa spec_sens no_omission prevalence equal_sens_spec
## thresholds 0.5518494 0.6987654   0.1003964  0.7704794       0.7466212
##            sensitivity
## thresholds   0.7019333