get_predict {Mapinguari}

Creates vectorized predict functions from models.

Description

get_predict Takes inputed models and create vectorized functions able to get the model predictionsfor any value inputed. Also outputs a table comparing models. nls gam glm lm randomForest gbm gls bam

Usage

get_predict(models, separator = "_", ...)

Arguments

models

list. List with models to create the prediction function. The model objects must have methods for function 'predict'.

separator

character. Character that separates variable names, years and scenarios.

...

additional arguments to be passed to predict function (specific for the method of the models supplied).

Value

Returns a list of vectorized functions that get predictions for the models inputted. The functions generated do not perform lazy evaluation, the user must be explicit

Examples

library(mgcv)

perf_no_size <-
  gamm(performance ~ s(temp, bs = 'cs'),
       random = list(id = ~ 1),
       data = TtorquatusPerformance)

perf_size <-
  gamm(performance ~ s(temp, bs = 'cs') + size,
       random = list(id = ~ 1),
       data = TtorquatusPerformance)

perf_functions <- get_predict(list(perf_s = perf_size,
                                   perf_ns = perf_no_size),
                              type = "response")

perf_nsFUN <- perf_functions$perf_ns
perf_sFUN <- perf_functions$perf_s

perf_nsFUN(temp = 30)
perf_sFUN(temp = 30, size = 70)
perf_nsFUN(temp = 30:35)
perf_sFUN(temp = 30, size = 70:75)
perf_sFUN(temp = 30:35, size = 70:75)