Create Model fit
create_fit.Rd
Create Model fit
Arguments
- model_recipe
a model recipe based on the predictors and target variable
- model_spec
a model specification that will eventually be fitted onto train/test data
- df
a dataframe that preferably consists ONLY the columns relevant to the entire regression model i.e, target variable and predictors
Examples
# Load data
data(mtcars)
# Create a target dataset
target_df <- target_df(mtcars, "gear", "wt", "qsec")
# Create recipe
model_recipe <- create_recipe(target_df, "gear")
# Create model specification with kmin
model_list <- list("mpg", "cyl", "disp", "hp", "am")
model_spec_kknn <- create_spec_kmin(target_df, model_recipe, "kknn", kmin=5, target_variable="gear")
model_spec_lm <- create_spec_kmin(target_df, model_recipe, "lm", target_variable="gear")
# Get first item from model_spec_kknn list
model_spec <- get_list_item(model_spec_kknn, 1)
# Fit model using kknn
model_fit_kknn <- create_fit(model_recipe, model_spec, target_df)
# Get first item from model_spec_lm list
model_spec <- get_list_item(model_spec_lm, 1)
# Fit model using lm
model_fit_lm <- create_fit(model_recipe, model_spec, target_df)