数据探索

children是结果变量，是一个2分类变量。

  1. library(tidyverse)
  2. ## -- Attaching packages ----------------------------------------------------------- tidyverse 1.3.1 --
  3. ## v ggplot2 3.3.3     v purrr   0.3.4
  4. ## v tibble  3.1.1     v dplyr   1.0.6
  5. ## v tidyr   1.1.3     v stringr 1.4.0
  6. ## v readr   1.4.0     v forcats 0.5.1
  7. ## -- Conflicts -------------------------------------------------------------- tidyverse_conflicts() --
  8. ## x dplyr::filter() masks stats::filter()
  9. ## x dplyr::lag()    masks stats::lag()
  11. hotels <- readr::read_csv("../datasets/tidytuesday/data/2020/2020-02-11/hotels.csv")
  12. ## 
  13. ## -- Column specification ----------------------------------------------------------------------------
  14. ## cols(
  15. ##   .default = col_double(),
  16. ##   hotel = col_character(),
  17. ##   arrival_date_month = col_character(),
  18. ##   meal = col_character(),
  19. ##   country = col_character(),
  20. ##   market_segment = col_character(),
  21. ##   distribution_channel = col_character(),
  22. ##   reserved_room_type = col_character(),
  23. ##   assigned_room_type = col_character(),
  24. ##   deposit_type = col_character(),
  25. ##   agent = col_character(),
  26. ##   company = col_character(),
  27. ##   customer_type = col_character(),
  28. ##   reservation_status = col_character(),
  29. ##   reservation_status_date = col_date(format = "")
  30. ## )
  31. ## i<U+00A0>Use `spec()` for the full column specifications.
  34. hotel_stays <- hotels %>%
  35.   filter(is_canceled == 0) %>%
  36.   mutate(
  37.     children = case_when( # 学习下case_when函数
  38.       children + babies > 0 ~ "children",
  39.       TRUE ~ "none"
  40.     ),
  41.     required_car_parking_spaces = case_when(
  42.       required_car_parking_spaces > 0 ~ "parking",
  43.       TRUE ~ "none"
  44.     )
  45.   ) %>%
  46.   select(-is_canceled, -reservation_status, -babies)
  48. hotel_stays
  49. ## # A tibble: 75,166 x 29
  50. ##    hotel      lead_time arrival_date_year arrival_date_mon~ arrival_date_week_n~
  51. ##    <chr>          <dbl>             <dbl> <chr>                            <dbl>
  52. ##  1 Resort Ho~       342              2015 July                                27
  53. ##  2 Resort Ho~       737              2015 July                                27
  54. ##  3 Resort Ho~         7              2015 July                                27
  55. ##  4 Resort Ho~        13              2015 July                                27
  56. ##  5 Resort Ho~        14              2015 July                                27
  57. ##  6 Resort Ho~        14              2015 July                                27
  58. ##  7 Resort Ho~         0              2015 July                                27
  59. ##  8 Resort Ho~         9              2015 July                                27
  60. ##  9 Resort Ho~        35              2015 July                                27
  61. ## 10 Resort Ho~        68              2015 July                                27
  62. ## # ... with 75,156 more rows, and 24 more variables:
  63. ## #   arrival_date_day_of_month <dbl>, stays_in_weekend_nights <dbl>,
  64. ## #   stays_in_week_nights <dbl>, adults <dbl>, children <chr>, meal <chr>,
  65. ## #   country <chr>, market_segment <chr>, distribution_channel <chr>,
  66. ## #   is_repeated_guest <dbl>, previous_cancellations <dbl>,
  67. ## #   previous_bookings_not_canceled <dbl>, reserved_room_type <chr>,
  68. ## #   assigned_room_type <chr>, booking_changes <dbl>, deposit_type <chr>,
  69. ## #   agent <chr>, company <chr>, days_in_waiting_list <dbl>,
  70. ## #   customer_type <chr>, adr <dbl>, required_car_parking_spaces <chr>,
  71. ## #   total_of_special_requests <dbl>, reservation_status_date <date>
  1. hotel_stays %>% count(children) # children是结果变量
  2. ## # A tibble: 2 x 2
  3. ##   children     n
  4. ##   <chr>    <int>
  5. ## 1 children  6073
  6. ## 2 none     69093
  1. skimr::skim(hotel_stays)

Table: Data summary
Name hotel_stays
Number of rows 75166
Number of columns 29
_
Column type frequency:
character 14
Date 1
numeric 14
__
Group variables None

Variable type: character

skim_variable n_missing complete_rate min max empty n_unique whitespace
hotel 0 1 10 12 0 2 0
arrival_date_month 0 1 3 9 0 12 0
children 0 1 4 8 0 2 0
meal 0 1 2 9 0 5 0
country 0 1 2 4 0 166 0
market_segment 0 1 6 13 0 7 0
distribution_channel 0 1 3 9 0 5 0
reserved_room_type 0 1 1 1 0 9 0
assigned_room_type 0 1 1 1 0 10 0
deposit_type 0 1 10 10 0 3 0
agent 0 1 1 4 0 315 0
company 0 1 1 4 0 332 0
customer_type 0 1 5 15 0 4 0
required_car_parking_spaces 0 1 4 7 0 2 0

Variable type: Date

skim_variable n_missing complete_rate min max median n_unique
reservation_status_date 0 1 2015-07-01 2017-09-14 2016-09-01 805

Variable type: numeric

skim_variable n_missing complete_rate mean sd p0 p25 p50 p75 p100 hist
lead_time 0 1 79.98 91.11 0.00 9.0 45.0 124 737 ▇▂▁▁▁
arrival_date_year 0 1 2016.15 0.70 2015.00 2016.0 2016.0 2017 2017 ▃▁▇▁▆
arrival_date_week_number 0 1 27.08 13.90 1.00 16.0 28.0 38 53 ▆▇▇▇▆
arrival_date_day_of_month 0 1 15.84 8.78 1.00 8.0 16.0 23 31 ▇▇▇▇▆
stays_in_weekend_nights 0 1 0.93 0.99 0.00 0.0 1.0 2 19 ▇▁▁▁▁
stays_in_week_nights 0 1 2.46 1.92 0.00 1.0 2.0 3 50 ▇▁▁▁▁
adults 0 1 1.83 0.51 0.00 2.0 2.0 2 4 ▁▂▇▁▁
is_repeated_guest 0 1 0.04 0.20 0.00 0.0 0.0 0 1 ▇▁▁▁▁
previous_cancellations 0 1 0.02 0.27 0.00 0.0 0.0 0 13 ▇▁▁▁▁
previous_bookings_not_canceled 0 1 0.20 1.81 0.00 0.0 0.0 0 72 ▇▁▁▁▁
booking_changes 0 1 0.29 0.74 0.00 0.0 0.0 0 21 ▇▁▁▁▁
days_in_waiting_list 0 1 1.59 14.78 0.00 0.0 0.0 0 379 ▇▁▁▁▁
adr 0 1 99.99 49.21 -6.38 67.5 92.5 125 510 ▇▆▁▁▁
total_of_special_requests 0 1 0.71 0.83 0.00 0.0 1.0 1 5 ▇▁▁▁▁ 
  1. # 写出这么刘畅的代码取决于对数据的理解以及对tidyverse包的理解
  2. hotel_stays %>%
  3.   mutate(arrival_date_month = factor(arrival_date_month, levels = month.name)) %>%
  4.   count(hotel, arrival_date_month, children) %>%
  5.   group_by(hotel, children) %>%
  6.   mutate(proportion = n / sum(n)) %>%
  7.   ggplot(aes(arrival_date_month, proportion, fill = children)) +
  8.   geom_col(position = "dodge") +
  9.   scale_y_continuous(labels = scales::percent_format()) + # 学习下百分比的用法
  10.   facet_wrap(~hotel, nrow = 2) +
  11.   labs(x = NULL, y = "Proportion of hotel stays", fill = NULL)

tidymodels-exercise-02 - 图1

  1. hotel_stays %>%
  2.   count(hotel, required_car_parking_spaces, children) %>%
  3.   group_by(hotel, children) %>%
  4.   mutate(proportion = n / sum(n)) %>%
  5.   ggplot(aes(required_car_parking_spaces, proportion, fill = children)) +
  6.   geom_col(position = "dodge") +
  7.   scale_y_continuous(labels = scales::percent_format()) +
  8.   facet_wrap(~hotel, nrow = 2) +
  9.   labs(x = NULL, y = "Proportion of hotel stays", fill = NULL)

tidymodels-exercise-02 - 图2

  1. library(GGally)
  2. ## Registered S3 method overwritten by 'GGally':
  3. ##   method from   
  4. ##   +.gg   ggplot2
  6. hotel_stays %>%
  7.   select(
  8.     children, adr,
  9.     required_car_parking_spaces,
  10.     total_of_special_requests
  11.   ) %>%
  12.   ggpairs(mapping = aes(color = children))
  13. ## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
  14. ## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
  15. ## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
  16. ## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

tidymodels-exercise-02 - 图3

建模

数据准备

  1. hotels_df <- hotel_stays %>%
  2.   select(
  3.     children, hotel, arrival_date_month, meal, adr, adults,
  4.     required_car_parking_spaces, total_of_special_requests,
  5.     stays_in_week_nights, stays_in_weekend_nights
  6.   ) %>%
  7.   mutate_if(is.character, factor)
  9. hotels_df <- hotels_df[1:1000,]
  1. library(tidymodels)
  2. ## -- Attaching packages ---------------------------------------------------------- tidymodels 0.1.3 --
  3. ## v broom        0.7.6      v rsample      0.0.9 
  4. ## v dials        0.0.9      v tune         0.1.5 
  5. ## v infer        0.5.4      v workflows    0.2.2 
  6. ## v modeldata    0.1.0      v workflowsets 0.0.2 
  7. ## v parsnip      0.1.5      v yardstick    0.0.8 
  8. ## v recipes      0.1.16
  9. ## -- Conflicts ------------------------------------------------------------- tidymodels_conflicts() --
  10. ## x scales::discard() masks purrr::discard()
  11. ## x dplyr::filter()   masks stats::filter()
  12. ## x recipes::fixed()  masks stringr::fixed()
  13. ## x dplyr::lag()      masks stats::lag()
  14. ## x yardstick::spec() masks readr::spec()
  15. ## x recipes::step()   masks stats::step()
  16. ## * Use tidymodels_prefer() to resolve common conflicts.
  17. tidymodels_prefer()
  19. set.seed(12)
  20. hotel_split <- hotels_df %>% 
  21.   initial_split()
  23. hotel_train <- training(hotel_split)
  24. hotel_test <- testing(hotel_split)
  26. hotel_rec <- recipe(children ~ ., data = hotel_train) %>%
  27.   themis::step_downsample(children) %>%
  28.   step_dummy(all_nominal(), -all_outcomes()) %>%
  29.   step_zv(all_numeric()) %>%
  30.   step_normalize(all_numeric()) %>%
  31.   prep()
  32. ## Registered S3 methods overwritten by 'themis':
  33. ##   method                  from   
  34. ##   bake.step_downsample    recipes
  35. ##   bake.step_upsample      recipes
  36. ##   prep.step_downsample    recipes
  37. ##   prep.step_upsample      recipes
  38. ##   tidy.step_downsample    recipes
  39. ##   tidy.step_upsample      recipes
  40. ##   tunable.step_downsample recipes
  41. ##   tunable.step_upsample   recipes
  43. hotel_rec
  44. ## Data Recipe
  45. ## 
  46. ## Inputs:
  47. ## 
  48. ##       role #variables
  49. ##    outcome          1
  50. ##  predictor          9
  51. ## 
  52. ## Training data contained 750 data points and no missing data.
  53. ## 
  54. ## Operations:
  55. ## 
  56. ## Down-sampling based on children [trained]
  57. ## Dummy variables from hotel, arrival_date_month, ... [trained]
  58. ## Zero variance filter removed 11 items [trained]
  59. ## Centering and scaling for adr, adults, ... [trained]
  1. # 提取准备好的训练集和测试集
  2. train_proc <- bake(hotel_rec, new_data = NULL)
  3. test_proc <- bake(hotel_rec, new_data = hotel_test)

knn

  1. knn_spec <- nearest_neighbor(mode = "classification") %>% set_engine("kknn")
  2. knn_fit <- knn_spec %>% fit(children ~ ., data = train_proc)
  4. knn_fit
  5. ## parsnip model object
  6. ## 
  7. ## Fit time:  11ms 
  8. ## 
  9. ## Call:
  10. ## kknn::train.kknn(formula = children ~ ., data = data, ks = min_rows(5,     data, 5))
  11. ## 
  12. ## Type of response variable: nominal
  13. ## Minimal misclassification: 0.3482143
  14. ## Best kernel: optimal
  15. ## Best k: 5

随机森林

  1. rf_spec <- rand_forest(mode = "classification") %>% set_engine("ranger")
  2. rf_fit <- rf_spec %>% fit(children ~ ., data = train_proc)
  4. rf_fit
  5. ## parsnip model object
  6. ## 
  7. ## Fit time:  80ms 
  8. ## Ranger result
  9. ## 
  10. ## Call:
  11. ##  ranger::ranger(x = maybe_data_frame(x), y = y, num.threads = 1,      verbose = FALSE, seed = sample.int(10^5, 1), probability = TRUE) 
  12. ## 
  13. ## Type:                             Probability estimation 
  14. ## Number of trees:                  500 
  15. ## Sample size:                      224 
  16. ## Number of independent variables:  11 
  17. ## Mtry:                             3 
  18. ## Target node size:                 10 
  19. ## Variable importance mode:         none 
  20. ## Splitrule:                        gini 
  21. ## OOB prediction error (Brier s.):  0.1935183

决策树

  1. tree_spec <- decision_tree(mode = "classification") %>% set_engine("rpart")
  2. tree_fit <- tree_spec %>% fit(children ~ ., data = train_proc)
  4. tree_fit
  5. ## parsnip model object
  6. ## 
  7. ## Fit time:  0ms 
  8. ## n= 224 
  9. ## 
  10. ## node), split, n, loss, yval, (yprob)
  11. ##       * denotes terminal node
  12. ## 
  13. ##   1) root 224 112 children (0.50000000 0.50000000)  
  14. ##     2) adr>=0.5847452 63  11 children (0.82539683 0.17460317) *
  15. ##     3) adr< 0.5847452 161  60 none (0.37267081 0.62732919)  
  16. ##       6) total_of_special_requests>=0.472056 43  15 children (0.65116279 0.34883721)  
  17. ##        12) meal_HB>=0.5359649 13   2 children (0.84615385 0.15384615) *
  18. ##        13) meal_HB< 0.5359649 30  13 children (0.56666667 0.43333333)  
  19. ##          26) adr< -0.266918 21   6 children (0.71428571 0.28571429) *
  20. ##          27) adr>=-0.266918 9   2 none (0.22222222 0.77777778) *
  21. ##       7) total_of_special_requests< 0.472056 118  32 none (0.27118644 0.72881356)  
  22. ##        14) adr>=-0.7217095 76  29 none (0.38157895 0.61842105)  
  23. ##          28) adr< -0.53929 8   2 children (0.75000000 0.25000000) *
  24. ##          29) adr>=-0.53929 68  23 none (0.33823529 0.66176471)  
  25. ##            58) adr>=-0.3662756 57  23 none (0.40350877 0.59649123)  
  26. ##             116) adr< 0.2846238 38  19 children (0.50000000 0.50000000)  
  27. ##               232) stays_in_weekend_nights>=-0.8218309 24   9 children (0.62500000 0.37500000) *
  28. ##               233) stays_in_weekend_nights< -0.8218309 14   4 none (0.28571429 0.71428571) *
  29. ##             117) adr>=0.2846238 19   4 none (0.21052632 0.78947368) *
  30. ##            59) adr< -0.3662756 11   0 none (0.00000000 1.00000000) *
  31. ##        15) adr< -0.7217095 42   3 none (0.07142857 0.92857143) *

评价模型

  1. set.seed(1234)
  3. # 模特卡洛法交叉验证
  4. validation_split <- mc_cv(data = train_proc, prop = 0.9, strata = children)
  5. validation_split
  6. ## # Monte Carlo cross-validation (0.9/0.1) with 25 resamples  using stratification 
  7. ## # A tibble: 25 x 2
  8. ##    splits           id        
  9. ##    <list>           <chr>     
  10. ##  1 <split [202/22]> Resample01
  11. ##  2 <split [202/22]> Resample02
  12. ##  3 <split [202/22]> Resample03
  13. ##  4 <split [202/22]> Resample04
  14. ##  5 <split [202/22]> Resample05
  15. ##  6 <split [202/22]> Resample06
  16. ##  7 <split [202/22]> Resample07
  17. ##  8 <split [202/22]> Resample08
  18. ##  9 <split [202/22]> Resample09
  19. ## 10 <split [202/22]> Resample10
  20. ## # ... with 15 more rows
  22. validation_split$splits[[1]] %>% analysis() %>% dim()
  23. ## [1] 202  12
  24. validation_split$splits[[1]] %>% assessment()
  25. ## # A tibble: 22 x 12
  26. ##       adr  adults total_of_special_~ stays_in_week_n~ stays_in_weekend~ children
  27. ##     <dbl>   <dbl>              <dbl>            <dbl>             <dbl> <fct>   
  28. ##  1 -0.909 -0.0643             0.957             0.675             0.671 children
  29. ##  2 -0.961 -2.94               0.957             0.193             0.671 children
  30. ##  3 -1.16  -0.0643             0.957            -0.288             0.671 children
  31. ##  4  0.837  2.82              -0.0130            0.193            -1.32  children
  32. ##  5 -0.698 -0.0643            -0.983             1.64              2.66  children
  33. ##  6  0.837 -0.0643            -0.0130           -1.73              0.671 children
  34. ##  7 -0.232 -0.0643            -0.0130           -0.770            -1.32  children
  35. ##  8  1.23  -0.0643            -0.983             0.675             2.66  children
  36. ##  9  1.09  -0.0643            -0.983             2.12              0.671 children
  37. ## 10  0.613 -0.0643             0.957            -0.770            -1.32  children
  38. ## # ... with 12 more rows, and 6 more variables: arrival_date_month_August <dbl>,
  39. ## #   arrival_date_month_July <dbl>, arrival_date_month_September <dbl>,
  40. ## #   meal_FB <dbl>, meal_HB <dbl>, required_car_parking_spaces_parking <dbl>
  1. knn_res <- fit_resamples(
  2.   knn_spec,
  3.   children ~ ., 
  4.   validation_split,
  5.   control = control_resamples(save_pred = T)
  6. )
  8. knn_res
  9. ## # Resampling results
  10. ## # Monte Carlo cross-validation (0.9/0.1) with 25 resamples  using stratification 
  11. ## # A tibble: 25 x 5
  12. ##    splits         id        .metrics         .notes           .predictions      
  13. ##    <list>         <chr>     <list>           <list>           <list>            
  14. ##  1 <split [202/2~ Resample~ <tibble[,4] [2 ~ <tibble[,1] [0 ~ <tibble[,6] [22 x~
  15. ##  2 <split [202/2~ Resample~ <tibble[,4] [2 ~ <tibble[,1] [0 ~ <tibble[,6] [22 x~
  16. ##  3 <split [202/2~ Resample~ <tibble[,4] [2 ~ <tibble[,1] [0 ~ <tibble[,6] [22 x~
  17. ##  4 <split [202/2~ Resample~ <tibble[,4] [2 ~ <tibble[,1] [0 ~ <tibble[,6] [22 x~
  18. ##  5 <split [202/2~ Resample~ <tibble[,4] [2 ~ <tibble[,1] [0 ~ <tibble[,6] [22 x~
  19. ##  6 <split [202/2~ Resample~ <tibble[,4] [2 ~ <tibble[,1] [0 ~ <tibble[,6] [22 x~
  20. ##  7 <split [202/2~ Resample~ <tibble[,4] [2 ~ <tibble[,1] [0 ~ <tibble[,6] [22 x~
  21. ##  8 <split [202/2~ Resample~ <tibble[,4] [2 ~ <tibble[,1] [0 ~ <tibble[,6] [22 x~
  22. ##  9 <split [202/2~ Resample~ <tibble[,4] [2 ~ <tibble[,1] [0 ~ <tibble[,6] [22 x~
  23. ## 10 <split [202/2~ Resample~ <tibble[,4] [2 ~ <tibble[,1] [0 ~ <tibble[,6] [22 x~
  24. ## # ... with 15 more rows
  1. knn_res %>% collect_metrics()
  2. ## # A tibble: 2 x 6
  3. ##   .metric  .estimator  mean     n std_err .config             
  4. ##   <chr>    <chr>      <dbl> <int>   <dbl> <chr>               
  5. ## 1 accuracy binary     0.649    25  0.0230 Preprocessor1_Model1
  6. ## 2 roc_auc  binary     0.646    25  0.0270 Preprocessor1_Model1
  7. knn_res %>% collect_metrics(summarise=FALSE)
  8. ## # A tibble: 2 x 6
  9. ##   .metric  .estimator  mean     n std_err .config             
  10. ##   <chr>    <chr>      <dbl> <int>   <dbl> <chr>               
  11. ## 1 accuracy binary     0.649    25  0.0230 Preprocessor1_Model1
  12. ## 2 roc_auc  binary     0.646    25  0.0270 Preprocessor1_Model1
  1. knn_res %>% collect_predictions()
  2. ## # A tibble: 550 x 7
  3. ##    id        .pred_children .pred_none  .row .pred_class children .config       
  4. ##    <chr>              <dbl>      <dbl> <int> <fct>       <fct>    <chr>         
  5. ##  1 Resample~         0.0822     0.918      4 none        children Preprocessor1~
  6. ##  2 Resample~         0.0822     0.918      5 none        children Preprocessor1~
  7. ##  3 Resample~         0.155      0.845      9 none        children Preprocessor1~
  8. ##  4 Resample~         0.561      0.439     16 children    children Preprocessor1~
  9. ##  5 Resample~         0          1         22 none        children Preprocessor1~
  10. ##  6 Resample~         0.180      0.820     28 none        children Preprocessor1~
  11. ##  7 Resample~         0.738      0.262     38 children    children Preprocessor1~
  12. ##  8 Resample~         0.845      0.155     80 children    children Preprocessor1~
  13. ##  9 Resample~         0.975      0.0251    98 children    children Preprocessor1~
  14. ## 10 Resample~         0          1        101 none        children Preprocessor1~
  15. ## # ... with 540 more rows
  1. rf_res <- fit_resamples(
  2.   rf_spec,
  3.   children ~ ., 
  4.   validation_split,
  5.   control = control_resamples(save_pred = T)
  6. )
  8. rf_res %>% collect_metrics()
  9. ## # A tibble: 2 x 6
  10. ##   .metric  .estimator  mean     n std_err .config             
  11. ##   <chr>    <chr>      <dbl> <int>   <dbl> <chr>               
  12. ## 1 accuracy binary     0.687    25  0.0158 Preprocessor1_Model1
  13. ## 2 roc_auc  binary     0.758    25  0.0154 Preprocessor1_Model1
  1. tree_res <- fit_resamples(
  2.   tree_spec,
  3.   children ~ ., 
  4.   validation_split,
  5.   control = control_resamples(save_pred = T)
  6. )
  8. tree_res %>% collect_metrics()
  9. ## # A tibble: 2 x 6
  10. ##   .metric  .estimator  mean     n std_err .config             
  11. ##   <chr>    <chr>      <dbl> <int>   <dbl> <chr>               
  12. ## 1 accuracy binary     0.713    25  0.0148 Preprocessor1_Model1
  13. ## 2 roc_auc  binary     0.752    25  0.0185 Preprocessor1_Model1
  1. knn_res %>%
  2.   unnest(.predictions) %>%
  3.   mutate(model = "kknn") %>%
  4.   bind_rows(tree_res %>%
  5.     unnest(.predictions) %>%
  6.     mutate(model = "rpart")) %>%
  7.   bind_rows(rf_res %>% 
  8.     unnest(.predictions) %>% 
  9.     mutate(model = "ranger")) %>% 
  10.   group_by(model) %>%
  11.   roc_curve(children, .pred_children) %>%
  12.   ggplot(aes(x = 1 - specificity, y = sensitivity, color = model)) +
  13.   geom_line(size = 1.5) +
  14.   geom_abline(
  15.     lty = 2, alpha = 0.5,
  16.     color = "gray50",
  17.     size = 1.2
  18.   )

tidymodels-exercise-02 - 图4

随机森林最好

  1. knn_conf <- knn_res %>% unnest(.predictions) %>% 
  2.   conf_mat(children, .pred_class)
  4. knn_conf
  5. ##           Truth
  6. ## Prediction children none
  7. ##   children      172   90
  8. ##   none          103  185
  1. knn_conf %>% autoplot()

tidymodels-exercise-02 - 图5

  1. rf_fit %>% predict(new_data = test_proc, type = "prob") %>% 
  2.   mutate(truth = hotel_test$children) %>% 
  3.   roc_auc(truth, .pred_children)
  4. ## # A tibble: 1 x 3
  5. ##   .metric .estimator .estimate
  6. ##   <chr>   <chr>          <dbl>
  7. ## 1 roc_auc binary         0.798