library(tidyverse)## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr 1.1.4 ✔ readr 2.1.5
## ✔ forcats 1.0.0 ✔ stringr 1.5.1
## ✔ ggplot2 3.5.1 ✔ tibble 3.2.1
## ✔ lubridate 1.9.4 ✔ tidyr 1.3.1
## ✔ purrr 1.0.4
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()
## ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errorsattrition_raw_tbl <- read_csv("../00_data/WA_Fn-UseC_-HR-Employee-Attrition.csv")## Rows: 1470 Columns: 35
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: ","
## chr (9): Attrition, BusinessTravel, Department, EducationField, Gender, Job...
## dbl (26): Age, DailyRate, DistanceFromHome, Education, EmployeeCount, Employ...
##
## ℹ Use `spec()` to retrieve the full column specification for this data.
## ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.# If data is not sensitive:
attrition_raw_tbl %>% glimpse()## Rows: 1,470
## Columns: 35
## $ Age <dbl> 41, 49, 37, 33, 27, 32, 59, 30, 38, 36, 35, 2…
## $ Attrition <chr> "Yes", "No", "Yes", "No", "No", "No", "No", "…
## $ BusinessTravel <chr> "Travel_Rarely", "Travel_Frequently", "Travel…
## $ DailyRate <dbl> 1102, 279, 1373, 1392, 591, 1005, 1324, 1358,…
## $ Department <chr> "Sales", "Research & Development", "Research …
## $ DistanceFromHome <dbl> 1, 8, 2, 3, 2, 2, 3, 24, 23, 27, 16, 15, 26, …
## $ Education <dbl> 2, 1, 2, 4, 1, 2, 3, 1, 3, 3, 3, 2, 1, 2, 3, …
## $ EducationField <chr> "Life Sciences", "Life Sciences", "Other", "L…
## $ EmployeeCount <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, …
## $ EmployeeNumber <dbl> 1, 2, 4, 5, 7, 8, 10, 11, 12, 13, 14, 15, 16,…
## $ EnvironmentSatisfaction <dbl> 2, 3, 4, 4, 1, 4, 3, 4, 4, 3, 1, 4, 1, 2, 3, …
## $ Gender <chr> "Female", "Male", "Male", "Female", "Male", "…
## $ HourlyRate <dbl> 94, 61, 92, 56, 40, 79, 81, 67, 44, 94, 84, 4…
## $ JobInvolvement <dbl> 3, 2, 2, 3, 3, 3, 4, 3, 2, 3, 4, 2, 3, 3, 2, …
## $ JobLevel <dbl> 2, 2, 1, 1, 1, 1, 1, 1, 3, 2, 1, 2, 1, 1, 1, …
## $ JobRole <chr> "Sales Executive", "Research Scientist", "Lab…
## $ JobSatisfaction <dbl> 4, 2, 3, 3, 2, 4, 1, 3, 3, 3, 2, 3, 3, 4, 3, …
## $ MaritalStatus <chr> "Single", "Married", "Single", "Married", "Ma…
## $ MonthlyIncome <dbl> 5993, 5130, 2090, 2909, 3468, 3068, 2670, 269…
## $ MonthlyRate <dbl> 19479, 24907, 2396, 23159, 16632, 11864, 9964…
## $ NumCompaniesWorked <dbl> 8, 1, 6, 1, 9, 0, 4, 1, 0, 6, 0, 0, 1, 0, 5, …
## $ Over18 <chr> "Y", "Y", "Y", "Y", "Y", "Y", "Y", "Y", "Y", …
## $ OverTime <chr> "Yes", "No", "Yes", "Yes", "No", "No", "Yes",…
## $ PercentSalaryHike <dbl> 11, 23, 15, 11, 12, 13, 20, 22, 21, 13, 13, 1…
## $ PerformanceRating <dbl> 3, 4, 3, 3, 3, 3, 4, 4, 4, 3, 3, 3, 3, 3, 3, …
## $ RelationshipSatisfaction <dbl> 1, 4, 2, 3, 4, 3, 1, 2, 2, 2, 3, 4, 4, 3, 2, …
## $ StandardHours <dbl> 80, 80, 80, 80, 80, 80, 80, 80, 80, 80, 80, 8…
## $ StockOptionLevel <dbl> 0, 1, 0, 0, 1, 0, 3, 1, 0, 2, 1, 0, 1, 1, 0, …
## $ TotalWorkingYears <dbl> 8, 10, 7, 8, 6, 8, 12, 1, 10, 17, 6, 10, 5, 3…
## $ TrainingTimesLastYear <dbl> 0, 3, 3, 3, 3, 2, 3, 2, 2, 3, 5, 3, 1, 2, 4, …
## $ WorkLifeBalance <dbl> 1, 3, 3, 3, 3, 2, 2, 3, 3, 2, 3, 3, 2, 3, 3, …
## $ YearsAtCompany <dbl> 6, 10, 0, 8, 2, 7, 1, 1, 9, 7, 5, 9, 5, 2, 4,…
## $ YearsInCurrentRole <dbl> 4, 7, 0, 7, 2, 7, 0, 0, 7, 7, 4, 5, 2, 2, 2, …
## $ YearsSinceLastPromotion <dbl> 0, 1, 0, 3, 2, 3, 0, 0, 1, 7, 0, 0, 4, 1, 0, …
## $ YearsWithCurrManager <dbl> 5, 7, 0, 0, 2, 6, 0, 0, 8, 7, 3, 8, 3, 2, 3, …The goal is to help predict attrition for employees.
Please write R code to create a predictive model that predicts the probability of attrition.
# Load libraries
library(tidymodels)## ── Attaching packages ────────────────────────────────────── tidymodels 1.2.0 ──## ✔ broom 1.0.7 ✔ rsample 1.2.1
## ✔ dials 1.4.0 ✔ tune 1.2.1
## ✔ infer 1.0.7 ✔ workflows 1.1.4
## ✔ modeldata 1.4.0 ✔ workflowsets 1.1.0
## ✔ parsnip 1.3.0 ✔ yardstick 1.3.2
## ✔ recipes 1.1.1## ── Conflicts ───────────────────────────────────────── tidymodels_conflicts() ──
## ✖ scales::discard() masks purrr::discard()
## ✖ dplyr::filter() masks stats::filter()
## ✖ recipes::fixed() masks stringr::fixed()
## ✖ dplyr::lag() masks stats::lag()
## ✖ yardstick::spec() masks readr::spec()
## ✖ recipes::step() masks stats::step()
## • Learn how to get started at https://www.tidymodels.org/start/library(dplyr)
library(janitor)## Warning: package 'janitor' was built under R version 4.4.3##
## Attaching package: 'janitor'## The following objects are masked from 'package:stats':
##
## chisq.test, fisher.test# Step 1: Clean and prepare data
attrition_tbl <- attrition_raw_tbl %>%
clean_names() %>%
mutate(attrition = factor(attrition, levels = c("No", "Yes")))
# Step 2: Split data
set.seed(123)
attrition_split <- initial_split(attrition_tbl, prop = 0.8, strata = attrition)
attrition_train <- training(attrition_split)
attrition_test <- testing(attrition_split)
# Step 3: Preprocessing recipe
attrition_recipe <- recipe(attrition ~ ., data = attrition_train) %>%
update_role(employee_number, new_role = "ID") %>%
step_rm(employee_count, over18, standard_hours) %>%
step_zv(all_predictors()) %>%
step_dummy(all_nominal_predictors()) %>%
step_normalize(all_numeric_predictors())
# Step 4: Define logistic regression model
logistic_model <- logistic_reg(mode = "classification", engine = "glm")
# Step 5: Create workflow
attrition_wf <- workflow() %>%
add_model(logistic_model) %>%
add_recipe(attrition_recipe)
# Step 6: Fit model
attrition_fit <- attrition_wf %>%
fit(data = attrition_train)
# Step 7: Make predictions on test data
attrition_preds <- predict(attrition_fit, new_data = attrition_test, type = "prob") %>%
bind_cols(attrition_test %>% select(attrition)) %>%
bind_cols(predict(attrition_fit, new_data = attrition_test, type = "class"))
# Step 8: Evaluate performance
metrics(attrition_preds, truth = attrition, estimate = .pred_class)## # A tibble: 2 × 3
## .metric .estimator .estimate
## <chr> <chr> <dbl>
## 1 accuracy binary 0.871
## 2 kap binary 0.464roc_auc(attrition_preds, truth = attrition, .pred_Yes)## # A tibble: 1 × 3
## .metric .estimator .estimate
## <chr> <chr> <dbl>
## 1 roc_auc binary 0.112Please update the code to use tidymodels instead of caret and to use the h2o model instead of glmnet.
# Load libraries
library(tidymodels)
library(h2o)## Warning: package 'h2o' was built under R version 4.4.3##
## ----------------------------------------------------------------------
##
## Your next step is to start H2O:
## > h2o.init()
##
## For H2O package documentation, ask for help:
## > ??h2o
##
## After starting H2O, you can use the Web UI at http://localhost:54321
## For more information visit https://docs.h2o.ai
##
## ----------------------------------------------------------------------##
## Attaching package: 'h2o'## The following objects are masked from 'package:lubridate':
##
## day, hour, month, week, year## The following objects are masked from 'package:stats':
##
## cor, sd, var## The following objects are masked from 'package:base':
##
## %*%, %in%, &&, ||, apply, as.factor, as.numeric, colnames,
## colnames<-, ifelse, is.character, is.factor, is.numeric, log,
## log10, log1p, log2, round, signif, trunclibrary(janitor)
library(dplyr)
# Initialize H2O
h2o.init()## Connection successful!
##
## R is connected to the H2O cluster:
## H2O cluster uptime: 47 minutes 10 seconds
## H2O cluster timezone: America/New_York
## H2O data parsing timezone: UTC
## H2O cluster version: 3.44.0.3
## H2O cluster version age: 1 year, 4 months and 10 days
## H2O cluster name: H2O_started_from_R_sheac_ggl822
## H2O cluster total nodes: 1
## H2O cluster total memory: 1.42 GB
## H2O cluster total cores: 8
## H2O cluster allowed cores: 8
## H2O cluster healthy: TRUE
## H2O Connection ip: localhost
## H2O Connection port: 54321
## H2O Connection proxy: NA
## H2O Internal Security: FALSE
## R Version: R version 4.4.2 (2024-10-31 ucrt)## Warning in h2o.clusterInfo():
## Your H2O cluster version is (1 year, 4 months and 10 days) old. There may be a newer version available.
## Please download and install the latest version from: https://h2o-release.s3.amazonaws.com/h2o/latest_stable.html# Step 1: Read data
attrition_raw_tbl <- readr::read_csv("C:/Users/sheac/OneDrive/Documents/PSU_DAT3100/00_data/WA_Fn-UseC_-HR-Employee-Attrition.csv")## Rows: 1470 Columns: 35## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: ","
## chr (9): Attrition, BusinessTravel, Department, EducationField, Gender, Job...
## dbl (26): Age, DailyRate, DistanceFromHome, Education, EmployeeCount, Employ...
##
## ℹ Use `spec()` to retrieve the full column specification for this data.
## ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.# Step 2: Clean and prepare
attrition_tbl <- attrition_raw_tbl %>%
clean_names() %>%
mutate(attrition = factor(attrition, levels = c("No", "Yes")))
# Step 3: Split data
set.seed(123)
attrition_split <- initial_split(attrition_tbl, prop = 0.8, strata = attrition)
attrition_train <- training(attrition_split)
attrition_test <- testing(attrition_split)
# Step 4: Recipe
attrition_recipe <- recipe(attrition ~ ., data = attrition_train) %>%
update_role(employee_number, new_role = "ID") %>%
step_rm(employee_count, over18, standard_hours) %>%
step_zv(all_predictors()) %>%
step_dummy(all_nominal_predictors()) %>%
step_normalize(all_numeric_predictors())
# Step 5: Prep and juice the data
prepped_recipe <- prep(attrition_recipe)
train_juiced <- juice(prepped_recipe)
test_baked <- bake(prepped_recipe, new_data = attrition_test)
# Step 6: Convert to H2O frame
train_h2o <- as.h2o(train_juiced)## | | | 0% | |======================================================================| 100%test_h2o <- as.h2o(test_baked)## | | | 0% | |======================================================================| 100%# Step 7: H2O model training (AutoML or GBM as example)
automl_model <- h2o.automl(
x = setdiff(names(train_h2o), c("attrition", "employee_number")),
y = "attrition",
training_frame = train_h2o,
max_runtime_secs = 60,
balance_classes = TRUE,
seed = 123
)## | | | 0% | |== | 2%
## 21:21:06.814: AutoML: XGBoost is not available; skipping it. | |====== | 8% | |========= | 12% | |============== | 20% | |==================== | 28% | |===================== | 30%
## 21:21:24.23: GBM_grid_1_AutoML_6_20250430_212106 [GBM Grid Search] failed: java.lang.AssertionError: Coldata -0.759627153880504 out of range hourly_rate:-1.69101314961767-1.5933479932449601 step=0.16421805714313148 nbins=20 actNBins=20 isInt=0
## cnt=3.000000, [-1.691013 - -1.526795], mean/var= 0.10/ 0.00,
## cnt=1.000000, [-1.526795 - -1.362577], mean/var= 0.08/ 0.00,
## cnt=0.000000, [-1.362577 - -1.198359], mean/var= 0.00/ 0.00,
## cnt=1.000000, [-1.198359 - -1.034141], mean/var= -0.24/ 0.00,
## cnt=1.000000, [-1.034141 - -0.869923], mean/var= 0.07/ 0.00,
## cnt=0.000000, [-0.869923 - -0.705705], mean/var= 0.00/ 0.00,
## cnt=0.000000, [-0.705705 - -0.541487], mean/var= 0.00/ 0.00,
## cnt=0.000000, [-0.541487 - -0.377269], mean/var= 0.00/ 0.00,
## cnt=0.000000, [-0.377269 - -0.213051], mean/var= 0.00/ 0.00,
## cnt=4.000000, [-0.213051 - -0.048833], mean/var= 0.13/ 0.00,
## cnt=3.000000, [-0.048833 - 0.115385], mean/var= 0.10/ 0.00,
## cnt=3.000000, [0.115385 - 0.279604], mean/var= 0.07/ 0.00,
## cnt=0.000000, [0.279604 - 0.443822], mean/var= 0.00/ 0.00,
## cnt=0.000000, [0.443822 - 0.608040], mean/var= 0.00/ 0.00,
## cnt=3.000000, [0.608040 - 0.772258], mean/var= 0.08/ 0.00,
## cnt=3.000000, [0.772258 - 0.936476], mean/var= 0.07/ 0.00,
## cnt=0.000000, [0.936476 - 1.100694], mean/var= 0.00/ 0.00,
## cnt=0.000000, [1.100694 - 1.264912], mean/var= 0.00/ 0.00,
## cnt=0.000000, [1.264912 - 1.429130], mean/var= 0.00/ 0.00,
## cnt=2.000000, [1.429130 - 1.593348], mean/var= 0.08/ 0.00,
## | |============================== | 43% | |==================================== | 51% | |======================================= | 56% | |========================================== | 60%
## 21:21:42.80: Model `GBM_grid_1_AutoML_6_20250430_212106_model_1` has unexpectedly been deleted from H2O: ignoring the model and/or removing it from the leaderboard.
## 21:21:42.80: Model `GBM_grid_1_AutoML_6_20250430_212106_model_3` has unexpectedly been deleted from H2O: ignoring the model and/or removing it from the leaderboard.
## 21:21:42.80: Model `GBM_grid_1_AutoML_6_20250430_212106_model_2` has unexpectedly been deleted from H2O: ignoring the model and/or removing it from the leaderboard. | |================================================ | 69% | |===================================================== | 75% | |======================================================== | 80% | |============================================================= | 87% | |================================================================ | 92% | |==================================================================== | 98% | |======================================================================| 100%# Step 8: Make predictions
preds <- h2o.predict(automl_model@leader, test_h2o)## | | | 0% | |======================================================================| 100%preds_df <- as.data.frame(preds)
# Step 9: Combine with actuals
results <- bind_cols(
attrition = test_baked$attrition,
predicted_class = preds_df$predict,
prob_no = preds_df$No,
prob_yes = preds_df$Yes
)
# Accuracy and other metrics
results_metrics <- metrics(results, truth = attrition, estimate = predicted_class)
# AUC
auc_value <- roc_auc(results, truth = attrition, prob_yes)
# Print
print(results_metrics)## # A tibble: 2 × 3
## .metric .estimator .estimate
## <chr> <chr> <dbl>
## 1 accuracy binary 0.861
## 2 kap binary 0.547print(auc_value)## # A tibble: 1 × 3
## .metric .estimator .estimate
## <chr> <chr> <dbl>
## 1 roc_auc binary 0.0973Error in .h2o.doSafeREST(h2oRestApiVersion = h2oRestApiVersion, urlSuffix = page, :
# 1. Confirm H2O is running
h2o.init()## Connection successful!
##
## R is connected to the H2O cluster:
## H2O cluster uptime: 52 minutes 40 seconds
## H2O cluster timezone: America/New_York
## H2O data parsing timezone: UTC
## H2O cluster version: 3.44.0.3
## H2O cluster version age: 1 year, 4 months and 10 days
## H2O cluster name: H2O_started_from_R_sheac_ggl822
## H2O cluster total nodes: 1
## H2O cluster total memory: 1.74 GB
## H2O cluster total cores: 8
## H2O cluster allowed cores: 8
## H2O cluster healthy: TRUE
## H2O Connection ip: localhost
## H2O Connection port: 54321
## H2O Connection proxy: NA
## H2O Internal Security: FALSE
## R Version: R version 4.4.2 (2024-10-31 ucrt)## Warning in h2o.clusterInfo():
## Your H2O cluster version is (1 year, 4 months and 10 days) old. There may be a newer version available.
## Please download and install the latest version from: https://h2o-release.s3.amazonaws.com/h2o/latest_stable.html# 2. Check Java version
system("java -version")## [1] 0# 3. Inspect your H2O frame
h2o.describe(train_h2o)## Label Type Missing Zeros PosInf NegInf
## 1 age real 0 0 0 0
## 2 daily_rate real 0 0 0 0
## 3 distance_from_home real 0 0 0 0
## 4 education real 0 0 0 0
## 5 employee_number int 0 0 0 0
## 6 environment_satisfaction real 0 0 0 0
## 7 hourly_rate real 0 0 0 0
## 8 job_involvement real 0 0 0 0
## 9 job_level real 0 0 0 0
## 10 job_satisfaction real 0 0 0 0
## 11 monthly_income real 0 0 0 0
## 12 monthly_rate real 0 0 0 0
## 13 num_companies_worked real 0 0 0 0
## 14 percent_salary_hike real 0 0 0 0
## 15 performance_rating real 0 0 0 0
## 16 relationship_satisfaction real 0 0 0 0
## 17 stock_option_level real 0 0 0 0
## 18 total_working_years real 0 0 0 0
## 19 training_times_last_year real 0 0 0 0
## 20 work_life_balance real 0 0 0 0
## 21 years_at_company real 0 0 0 0
## 22 years_in_current_role real 0 0 0 0
## 23 years_since_last_promotion real 0 0 0 0
## 24 years_with_curr_manager real 0 0 0 0
## 25 attrition enum 0 986 0 0
## 26 business_travel_Travel_Frequently real 0 0 0 0
## 27 business_travel_Travel_Rarely real 0 0 0 0
## 28 department_Research...Development real 0 0 0 0
## 29 department_Sales real 0 0 0 0
## 30 education_field_Life.Sciences real 0 0 0 0
## 31 education_field_Marketing real 0 0 0 0
## 32 education_field_Medical real 0 0 0 0
## 33 education_field_Other real 0 0 0 0
## 34 education_field_Technical.Degree real 0 0 0 0
## 35 gender_Male real 0 0 0 0
## 36 job_role_Human.Resources real 0 0 0 0
## 37 job_role_Laboratory.Technician real 0 0 0 0
## 38 job_role_Manager real 0 0 0 0
## 39 job_role_Manufacturing.Director real 0 0 0 0
## 40 job_role_Research.Director real 0 0 0 0
## 41 job_role_Research.Scientist real 0 0 0 0
## 42 job_role_Sales.Executive real 0 0 0 0
## 43 job_role_Sales.Representative real 0 0 0 0
## 44 marital_status_Married real 0 0 0 0
## 45 marital_status_Single real 0 0 0 0
## 46 over_time_Yes real 0 0 0 0
## Min Max Mean Sigma Cardinality
## 1 -2.0850036 2.5710861 -2.558717e-17 1.00000 NA
## 2 -1.7378056 1.7296501 9.662952e-17 1.00000 NA
## 3 -1.0108389 2.4715276 -6.084543e-16 1.00000 NA
## 4 -1.8685813 2.0468588 -5.117434e-16 1.00000 NA
## 5 1.0000000 2068.0000000 1.023358e+03 605.62206 NA
## 6 -1.6089143 1.1482189 7.977560e-16 1.00000 NA
## 7 -1.7400335 1.6913886 -2.435118e-16 1.00000 NA
## 8 -2.4311929 1.7904598 -7.620315e-16 1.00000 NA
## 9 -0.9574690 2.6775103 3.563298e-16 1.00000 NA
## 10 -1.5787506 1.1532553 -6.608483e-16 1.00000 NA
## 11 -1.1693596 2.8802210 -4.384920e-17 1.00000 NA
## 12 -1.7184862 1.7900028 6.650056e-18 1.00000 NA
## 13 -1.0721293 2.5374750 1.973248e-17 1.00000 NA
## 14 -1.1512279 2.6461852 -1.139496e-15 1.00000 NA
## 15 -0.4362500 2.2903126 -7.685909e-16 1.00000 NA
## 16 -1.5782605 1.1923413 1.031727e-15 1.00000 NA
## 17 -0.9382129 2.5763466 2.818926e-17 1.00000 NA
## 18 -1.4425561 3.6945964 -6.093487e-16 1.00000 NA
## 19 -2.1663889 2.4815240 4.068198e-16 1.00000 NA
## 20 -2.5052487 1.7445115 1.880007e-16 1.00000 NA
## 21 -1.1471747 5.4305013 -1.369889e-16 1.00000 NA
## 22 -1.1785978 3.8480898 -1.099652e-15 1.00000 NA
## 23 -0.6897883 4.0315780 1.075529e-16 1.00000 NA
## 24 -1.1631352 3.6421647 -5.448116e-16 1.00000 NA
## 25 0.0000000 1.0000000 1.608511e-01 0.36755 2
## 26 -0.4730409 2.1121828 8.918647e-16 1.00000 NA
## 27 -1.5961450 0.6259763 1.127137e-15 1.00000 NA
## 28 -1.3756732 0.7262982 -1.172456e-15 1.00000 NA
## 29 -0.6523867 1.5315288 4.547144e-16 1.00000 NA
## 30 -0.8365632 1.1943496 -1.012645e-15 1.00000 NA
## 31 -0.3433402 2.9100847 -1.088539e-16 1.00000 NA
## 32 -0.6749976 1.4802258 9.384854e-16 1.00000 NA
## 33 -0.2572680 3.8836893 2.016074e-16 1.00000 NA
## 34 -0.3081877 3.2420140 2.321819e-16 1.00000 NA
## 35 -1.2177359 0.8204972 2.146395e-15 1.00000 NA
## 36 -0.2017656 4.9520289 -4.867797e-16 1.00000 NA
## 37 -0.4676466 2.1365468 -3.416321e-16 1.00000 NA
## 38 -0.2647033 3.7745999 4.653396e-16 1.00000 NA
## 39 -0.3324032 3.0058342 5.721877e-16 1.00000 NA
## 40 -0.2458004 4.0648787 6.936997e-16 1.00000 NA
## 41 -0.4917986 2.0316222 -1.097375e-15 1.00000 NA
## 42 -0.5341488 1.8705441 -1.298007e-15 1.00000 NA
## 43 -0.2379211 4.1994966 1.163078e-16 1.00000 NA
## 44 -0.9312925 1.0728626 7.286923e-16 1.00000 NA
## 45 -0.6763333 1.4773026 5.431853e-16 1.00000 NA
## 46 -0.6312449 1.5828230 3.150692e-16 1.00000 NAPlease update the code to use h2o.performance in Step 5, instead of mean.
# Load libraries
library(tidymodels)
library(h2o)
library(janitor)
library(dplyr)
# Initialize H2O
h2o.init()## Connection successful!
##
## R is connected to the H2O cluster:
## H2O cluster uptime: 52 minutes 44 seconds
## H2O cluster timezone: America/New_York
## H2O data parsing timezone: UTC
## H2O cluster version: 3.44.0.3
## H2O cluster version age: 1 year, 4 months and 10 days
## H2O cluster name: H2O_started_from_R_sheac_ggl822
## H2O cluster total nodes: 1
## H2O cluster total memory: 1.74 GB
## H2O cluster total cores: 8
## H2O cluster allowed cores: 8
## H2O cluster healthy: TRUE
## H2O Connection ip: localhost
## H2O Connection port: 54321
## H2O Connection proxy: NA
## H2O Internal Security: FALSE
## R Version: R version 4.4.2 (2024-10-31 ucrt)## Warning in h2o.clusterInfo():
## Your H2O cluster version is (1 year, 4 months and 10 days) old. There may be a newer version available.
## Please download and install the latest version from: https://h2o-release.s3.amazonaws.com/h2o/latest_stable.html# Step 1: Read and prepare data
attrition_raw_tbl <- readr::read_csv("C:/Users/sheac/OneDrive/Documents/PSU_DAT3100/00_data/WA_Fn-UseC_-HR-Employee-Attrition.csv")## Rows: 1470 Columns: 35
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: ","
## chr (9): Attrition, BusinessTravel, Department, EducationField, Gender, Job...
## dbl (26): Age, DailyRate, DistanceFromHome, Education, EmployeeCount, Employ...
##
## ℹ Use `spec()` to retrieve the full column specification for this data.
## ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.attrition_tbl <- attrition_raw_tbl %>%
clean_names() %>%
mutate(attrition = factor(attrition, levels = c("No", "Yes")))
# Step 2: Split data
set.seed(123)
attrition_split <- initial_split(attrition_tbl, prop = 0.8, strata = attrition)
attrition_train <- training(attrition_split)
attrition_test <- testing(attrition_split)
# Step 3: Preprocessing with recipe
attrition_recipe <- recipe(attrition ~ ., data = attrition_train) %>%
update_role(employee_number, new_role = "ID") %>%
step_rm(employee_count, over18, standard_hours) %>%
step_zv(all_predictors()) %>%
step_dummy(all_nominal_predictors()) %>%
step_normalize(all_numeric_predictors())
# Step 4: Prepare data for H2O
prepped_recipe <- prep(attrition_recipe)
train_juiced <- juice(prepped_recipe)
test_baked <- bake(prepped_recipe, new_data = attrition_test)
train_h2o <- as.h2o(train_juiced)## | | | 0% | |======================================================================| 100%test_h2o <- as.h2o(test_baked)## | | | 0% | |======================================================================| 100%# Set x and y
x <- setdiff(names(train_h2o), c("attrition", "employee_number"))
y <- "attrition"
# Step 5: Train model with H2O AutoML
automl_model <- h2o.automl(
x = x,
y = y,
training_frame = train_h2o,
max_runtime_secs = 60,
balance_classes = TRUE,
seed = 123
)## | | | 0% | |== | 2%
## 21:26:44.113: AutoML: XGBoost is not available; skipping it. | |====== | 8% | |========= | 12% | |============== | 19% | |==================== | 28% | |==================== | 29% | |===================== | 31% | |======================== | 35% | |============================ | 40% | |=============================== | 44% | |================================== | 49% | |====================================== | 54% | |========================================== | 59% | |============================================= | 64% | |=================================================== | 73% | |======================================================== | 80% | |============================================================= | 87% | |================================================================= | 92% | |==================================================================== | 98% | |======================================================================| 100%# Step 6: Evaluate performance using h2o.performance
perf <- h2o.performance(automl_model@leader, newdata = test_h2o)
# Print performance metrics
print(perf)## H2OBinomialMetrics: stackedensemble
##
## MSE: 0.0824071
## RMSE: 0.2870664
## LogLoss: 0.2760028
## Mean Per-Class Error: 0.1698296
## AUC: 0.9042679
## AUCPR: 0.7041058
## Gini: 0.8085358
##
## Confusion Matrix (vertical: actual; across: predicted) for F1-optimal threshold:
## No Yes Error Rate
## No 230 17 0.068826 =17/247
## Yes 13 35 0.270833 =13/48
## Totals 243 52 0.101695 =30/295
##
## Maximum Metrics: Maximum metrics at their respective thresholds
## metric threshold value idx
## 1 max f1 0.339026 0.700000 51
## 2 max f2 0.177814 0.742049 90
## 3 max f0point5 0.444626 0.706522 33
## 4 max accuracy 0.444626 0.898305 33
## 5 max precision 0.952680 1.000000 0
## 6 max recall 0.038098 1.000000 212
## 7 max specificity 0.952680 1.000000 0
## 8 max absolute_mcc 0.339026 0.639635 51
## 9 max min_per_class_accuracy 0.230490 0.833333 78
## 10 max mean_per_class_accuracy 0.177814 0.838310 90
## 11 max tns 0.952680 247.000000 0
## 12 max fns 0.952680 47.000000 0
## 13 max fps 0.000741 247.000000 294
## 14 max tps 0.038098 48.000000 212
## 15 max tnr 0.952680 1.000000 0
## 16 max fnr 0.952680 0.979167 0
## 17 max fpr 0.000741 1.000000 294
## 18 max tpr 0.038098 1.000000 212
##
## Gains/Lift Table: Extract with `h2o.gainsLift(<model>, <data>)` or `h2o.gainsLift(<model>, valid=<T/F>, xval=<T/F>)`h2o.auc(perf)## [1] 0.9042679h2o.confusionMatrix(perf)## Confusion Matrix (vertical: actual; across: predicted) for max f1 @ threshold = 0.339025778684586:
## No Yes Error Rate
## No 230 17 0.068826 =17/247
## Yes 13 35 0.270833 =13/48
## Totals 243 52 0.101695 =30/295h2o.logloss(perf)## [1] 0.2760028