second tutorial from tidymodels

02_preprocess_with_recipes

##################################################################
##                          tidymodels                         ##
##                  2 Preprocess with Recipes                   ##
##        url: https://www.tidymodels.org/start/recipes/        ##
##################################################################

# 1.0 INTRODUCTION ----
        ## 1.1 tidymodels ----
        library(tidymodels)
        ## 1.2 Helper packages
        library(nycflights13)    # for flight data
        library(skimr)           # for variable summaries
# 2.0 THE NEW YORK CITY FLIGHT DATA ----
        ## 2.1 Set seed ----
        set.seed(123)
        ## 2.2 Load ----
        data(flights)
        ## 2.3 View original ----
        skimr::skim(flights)
        ## 2.4 Change ----
        flight_data <-
                flights %>%
                mutate(
                        # Convert the arrival delay to a factor
                        arr_delay = ifelse(arr_delay >= 30, "late", "on_time"),
                        arr_delay = factor(arr_delay),
                        # We will use the date (not date-time) in the recipe below
                        date = as.Date(time_hour)
                ) %>%
                # Include the weather data
                inner_join(weather, by = c("origin", "time_hour")) %>%
                # Only retain the specific columns we will use
                select(dep_time, flight, origin, dest, air_time, distance,
                       carrier, date, arr_delay, time_hour) %>%
                # Exclude missing data
                na.omit() %>%
                # For creating models, it is better to have qualitative columns
                # encoded as factors (instead of character strings)
                mutate_if(is.character, as.factor)
        ##2.5 View changes -----
        glimpse(flight_data)

#3.0 DATA SPLITTING ----
        ## 3.1 Set seed ----
        set.seed(555)
        ## 3.2 Split ----
        data_split <- initial_split(flight_data, prop = 3/4)
        ## 3.3 Training & Testing ----
        train_data <- training(data_split)
        test_data  <- testing(data_split)

# 4.0 CREATE RECIPE AND ROLES ----
        ## 4.1 Original recipe ----
        flights_rec <-
        recipe(arr_delay ~ ., data = train_data)
        ## 4.2 New recipe ----
        #two columns designated as 'ID' and not variables
        flights_rec <-
                recipe(arr_delay ~ ., data = train_data) %>%
                update_role(flight, time_hour, new_role = "ID")
        summary(flights_rec)
# 5.0 FEATURE ENGINEERING ----
flights_rec <-
        recipe(arr_delay ~ ., data = train_data) %>%
        update_role(flight, time_hour, new_role = "ID") %>%
        step_date(date, features = c("dow", "month")) %>%
        step_holiday(date, holidays = timeDate::listHolidays("US")) %>%
        step_rm(date) %>%
        step_dummy(all_nominal(), -all_outcomes()) %>%
        step_zv(all_predictors())
# 6.0 FIT A MODEL WITH A RECIPE ----
        ## 6.1 Specify model ----
        lr_mod <-
                logistic_reg() %>%
                set_engine("glm")
        ## 6.2 Specify workflow ----
        flights_wflow <-
                workflow() %>%
                add_model(lr_mod) %>%
                add_recipe(flights_rec)
        flights_wflow
        ## 6.3 Fit model ----
        flights_fit <-
                flights_wflow %>%
                fit(data = train_data)
        ## 6.4 Extract results ----
        flights_fit %>%
                pull_workflow_fit() %>%
                tidy()
# 7.0 USE A TRAINED WORKFLOW TO PREDICT ----
        ## 7.1 Predict -- returns factor ----
        predict(flights_fit, test_data)
        ## 7.2 Predict -- return probability ----
        flights_pred <-
                predict(flights_fit, test_data, type = "prob") %>%
                bind_cols(test_data %>% select(arr_delay, time_hour, flight))
        flights_pred
        ## 7.3 Plot -- library('yardstick') ----
        flights_pred %>%
                roc_curve(truth = arr_delay, .pred_late) %>%
                autoplot()
        ## 7.4 Area under ROC curve ----
        flights_pred %>%
                roc_auc(truth = arr_delay, .pred_late)