# Import Data

library(tidyverse)
## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr     1.1.3     ✔ readr     2.1.4
## ✔ forcats   1.0.0     ✔ stringr   1.5.0
## ✔ ggplot2   3.4.3     ✔ tibble    3.2.1
## ✔ lubridate 1.9.3     ✔ tidyr     1.3.0
## ✔ purrr     1.0.2     
## ── 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 errors
data <- readr::read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2020/2020-09-22/members.csv')
## Rows: 76519 Columns: 21
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: ","
## chr (10): expedition_id, member_id, peak_id, peak_name, season, sex, citizen...
## dbl  (5): year, age, highpoint_metres, death_height_metres, injury_height_me...
## lgl  (6): hired, success, solo, oxygen_used, died, injured
## 
## 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.
data %>% skimr::skim()
Data summary
Name Piped data
Number of rows 76519
Number of columns 21
_______________________
Column type frequency:
character 10
logical 6
numeric 5
________________________
Group variables None

Variable type: character

skim_variable n_missing complete_rate min max empty n_unique whitespace
expedition_id 0 1.00 9 9 0 10350 0
member_id 0 1.00 12 12 0 76518 0
peak_id 0 1.00 4 4 0 391 0
peak_name 15 1.00 4 25 0 390 0
season 0 1.00 6 7 0 5 0
sex 2 1.00 1 1 0 2 0
citizenship 10 1.00 2 23 0 212 0
expedition_role 21 1.00 4 25 0 524 0
death_cause 75413 0.01 3 27 0 12 0
injury_type 74807 0.02 3 27 0 11 0

Variable type: logical

skim_variable n_missing complete_rate mean count
hired 0 1 0.21 FAL: 60788, TRU: 15731
success 0 1 0.38 FAL: 47320, TRU: 29199
solo 0 1 0.00 FAL: 76398, TRU: 121
oxygen_used 0 1 0.24 FAL: 58286, TRU: 18233
died 0 1 0.01 FAL: 75413, TRU: 1106
injured 0 1 0.02 FAL: 74806, TRU: 1713

Variable type: numeric

skim_variable n_missing complete_rate mean sd p0 p25 p50 p75 p100 hist
year 0 1.00 2000.36 14.78 1905 1991 2004 2012 2019 ▁▁▁▃▇
age 3497 0.95 37.33 10.40 7 29 36 44 85 ▁▇▅▁▁
highpoint_metres 21833 0.71 7470.68 1040.06 3800 6700 7400 8400 8850 ▁▁▆▃▇
death_height_metres 75451 0.01 6592.85 1308.19 400 5800 6600 7550 8830 ▁▁▂▇▆
injury_height_metres 75510 0.01 7049.91 1214.24 400 6200 7100 8000 8880 ▁▁▂▇▇

Goal: Predict the death on Nepal Himalaya climbers

Issues with data:

Cleaning Data

factors_vec <- data %>% select(year, age, highpoint_metres, death_height_metres, injury_height_metres) %>%
    names()

# Treating missing values
data_clean <- data %>% 
    select(-death_cause, -injury_type, -death_height_metres, -injury_height_metres) %>%
    na.omit() %>%
    
    # Mutate logical Variables
    mutate(died = case_when(died == "TRUE" ~ "died", died == "FALSE" ~ "no")) %>%
    
    mutate(across(where(is.logical), as.factor)) %>%

    # Recode "died"
    mutate(died = if_else(died == "TRUE", "deaths", died))

Explore Data

data_clean %>% count(died)
## # A tibble: 2 × 2
##   died      n
##   <chr> <int>
## 1 died    744
## 2 no    51639
data_clean %>%
    ggplot(aes(died)) +
    geom_bar()

Died vs. expedition_role

top_10_exp_role_vec <- data_clean %>% 
    count(expedition_role, sort = TRUE) %>% 
    head(10) %>% 
    pull(expedition_role)

# Relationship between pay and attrition
data_clean %>%
    filter(expedition_role %in% top_10_exp_role_vec) %>%
    count(died, expedition_role) %>%
    ggplot(aes(died, expedition_role, fill = n)) +
    geom_tile()

Relationship in all variables with Correlation plot

library(correlationfunnel)
## ══ correlationfunnel Tip #1 ════════════════════════════════════════════════════
## Make sure your data is not overly imbalanced prior to using `correlate()`.
## If less than 5% imbalance, consider sampling. :)
 # Step 1: Binarize
data_binarized <- data_clean %>%
    select(-member_id) %>% # ID variable
    binarize()

data_binarized %>% glimpse()
## Rows: 52,383
## Columns: 82
## $ expedition_id__HIML13308       <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ `expedition_id__-OTHER`        <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, …
## $ peak_id__AMAD                  <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, …
## $ peak_id__ANN1                  <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_id__BARU                  <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_id__CHOY                  <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_id__DHA1                  <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_id__EVER                  <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_id__HIML                  <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_id__KANG                  <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_id__LHOT                  <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_id__MAKA                  <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_id__MANA                  <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_id__PUMO                  <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ `peak_id__-OTHER`              <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_name__Ama_Dablam          <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, …
## $ peak_name__Annapurna_I         <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_name__Baruntse            <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_name__Cho_Oyu             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_name__Dhaulagiri_I        <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_name__Everest             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_name__Himlung_Himal       <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_name__Kangchenjunga       <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_name__Lhotse              <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_name__Makalu              <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_name__Manaslu             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ peak_name__Pumori              <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ `peak_name__-OTHER`            <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ `year__-Inf_1997`              <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, …
## $ year__1997_2007                <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ year__2007_2012                <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ year__2012_Inf                 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ season__Autumn                 <dbl> 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ season__Spring                 <dbl> 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, …
## $ season__Winter                 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ `season__-OTHER`               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ sex__F                         <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ sex__M                         <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, …
## $ `age__-Inf_29`                 <dbl> 0, 0, 1, 0, 1, 0, 1, 0, 0, 1, 0, 0, 1, …
## $ age__29_36                     <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 1, 0, …
## $ age__36_43                     <dbl> 1, 1, 0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, …
## $ age__43_Inf                    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ citizenship__Australia         <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ citizenship__Austria           <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ citizenship__Canada            <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ citizenship__China             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ citizenship__France            <dbl> 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ citizenship__Germany           <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ citizenship__India             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ citizenship__Italy             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ citizenship__Japan             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ citizenship__Nepal             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ citizenship__New_Zealand       <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ citizenship__Poland            <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ citizenship__Russia            <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ citizenship__S_Korea           <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ citizenship__Spain             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ citizenship__Switzerland       <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ citizenship__UK                <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ citizenship__USA               <dbl> 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, …
## $ `citizenship__-OTHER`          <dbl> 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, …
## $ expedition_role__Climber       <dbl> 0, 0, 1, 1, 1, 1, 1, 0, 0, 1, 0, 1, 0, …
## $ expedition_role__Deputy_Leader <dbl> 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ `expedition_role__H-A_Worker`  <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ expedition_role__Leader        <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ `expedition_role__-OTHER`      <dbl> 0, 1, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 1, …
## $ hired__FALSE                   <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, …
## $ hired__TRUE                    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ `highpoint_metres__-Inf_6750`  <dbl> 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ highpoint_metres__6750_7400    <dbl> 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, …
## $ highpoint_metres__7400_8450    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ highpoint_metres__8450_Inf     <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ success__FALSE                 <dbl> 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ success__TRUE                  <dbl> 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, …
## $ solo__FALSE                    <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, …
## $ `solo__-OTHER`                 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ oxygen_used__FALSE             <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, …
## $ oxygen_used__TRUE              <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ died__died                     <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ died__no                       <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, …
## $ injured__FALSE                 <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, …
## $ injured__TRUE                  <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
# Step 2: Correlation
data_correlation <- data_binarized %>%
    correlate(died__died)
## Warning: correlate(): [Data Imbalance Detected] Consider sampling to balance the classes more than 5%
##   Column with imbalance: died__died
data_correlation
## # A tibble: 82 × 3
##    feature   bin         correlation
##    <fct>     <chr>             <dbl>
##  1 died      died             1     
##  2 died      no              -1     
##  3 year      -Inf_1997        0.0843
##  4 success   FALSE            0.0562
##  5 success   TRUE            -0.0562
##  6 peak_id   ANN1             0.0431
##  7 peak_name Annapurna_I      0.0431
##  8 year      2012_Inf        -0.0330
##  9 peak_id   AMAD            -0.0323
## 10 peak_name Ama_Dablam      -0.0323
## # ℹ 72 more rows
# Step 3: Plot
data_correlation %>%
    correlationfunnel::plot_correlation_funnel()
## Warning: ggrepel: 45 unlabeled data points (too many overlaps). Consider
## increasing max.overlaps

Model Building

Split Data

library(tidymodels)
## ── Attaching packages ────────────────────────────────────── tidymodels 1.1.1 ──
## ✔ broom        1.0.5     ✔ rsample      1.2.0
## ✔ dials        1.2.0     ✔ tune         1.1.2
## ✔ infer        1.0.6     ✔ workflows    1.1.3
## ✔ modeldata    1.3.0     ✔ workflowsets 1.0.1
## ✔ parsnip      1.1.1     ✔ yardstick    1.3.0
## ✔ recipes      1.0.8
## ── 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()
## • Use suppressPackageStartupMessages() to eliminate package startup messages
set.seed(1234)
 data_clean <- data_clean %>% group_by(died) %>% sample_n(50) %>% ungroup()

data_split <- initial_split(data_clean, strata = died)
data_training <- training(data_split)
data_test <- testing(data_split)

 # Cross Validation
data_cv <- rsample::vfold_cv(data_training, strata = died)
data_cv
## #  10-fold cross-validation using stratification 
## # A tibble: 10 × 2
##    splits         id    
##    <list>         <chr> 
##  1 <split [66/8]> Fold01
##  2 <split [66/8]> Fold02
##  3 <split [66/8]> Fold03
##  4 <split [66/8]> Fold04
##  5 <split [66/8]> Fold05
##  6 <split [66/8]> Fold06
##  7 <split [66/8]> Fold07
##  8 <split [68/6]> Fold08
##  9 <split [68/6]> Fold09
## 10 <split [68/6]> Fold10

Preprocess Data

  # Solving the unbalanced target variable
library(themis)

xgboost_rec <- recipes::recipe(died ~ ., data = data_training) %>%
    update_role(member_id, new_role = "ID") %>%
    step_other(peak_name, citizenship, expedition_role) %>%
    step_dummy(all_nominal_predictors(), one_hot = TRUE)
    
    
    
xgboost_rec %>% prep() %>% juice() %>% glimpse()
## Rows: 74
## Columns: 136
## $ member_id                  <fct> DHA192101-11, MANA12109-01, LHOT12301-08, E…
## $ year                       <dbl> 1992, 2012, 2012, 1982, 1989, 1992, 2019, 2…
## $ age                        <dbl> 37, 34, 39, 31, 30, 42, 39, 32, 24, 23, 30,…
## $ highpoint_metres           <dbl> 7000, 8163, 8250, 8230, 7315, 6500, 7200, 8…
## $ died                       <fct> died, died, died, died, died, died, died, d…
## $ expedition_id_AMAD00323    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_AMAD92102    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0…
## $ expedition_id_AMPG59101    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_ANN109303    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_ANN385101    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_CHOY03101    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_CHOY03303    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_CHOY04303    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_CHOY14322    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_CHOY14324    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_CHOY19101    <dbl> 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_CHOY91106    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_CHRE88101    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0…
## $ expedition_id_DHA189402    <dbl> 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_DHA192101    <dbl> 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_DHA192102    <dbl> 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_DHA198101    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_DORJ04301    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER00303    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER05115    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER06120    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER07148    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER09179    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0…
## $ expedition_id_EVER10157    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER11103    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER11142    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER12101    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER12112    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER12173    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER12180    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER13133    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1…
## $ expedition_id_EVER15122    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER16134    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER17187    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER18125    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER19138    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER38101    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER74301    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER82103    <dbl> 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER82302    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER84102    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER85303    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER87102    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER87103    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_EVER97108    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_GAN485301    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_GURJ85101    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_HIME77101    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0…
## $ expedition_id_HIML14301    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_JANE98101    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_KANG13102    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_LANG64301    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_LANG89101    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_LHOT05102    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_LHOT12301    <dbl> 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_LSHR87301    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0…
## $ expedition_id_LSIS18301    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_MAKA11108    <dbl> 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_MANA05102    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_MANA08102    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_MANA09113    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_MANA12109    <dbl> 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_MANA12317    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_MANA82101    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_MANA89103    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_MANA91104    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_NANG10301    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_NPHU14301    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_NUPT75101    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0…
## $ expedition_id_PUMO05104    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_PUMO91301    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_RATC09301    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_id_TUKU70101    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_AMAD               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0…
## $ peak_id_AMPG               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_ANN1               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_ANN3               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_CHOY               <dbl> 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_CHRE               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0…
## $ peak_id_DHA1               <dbl> 1, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_DORJ               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_EVER               <dbl> 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1…
## $ peak_id_GAN4               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_GURJ               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_HIME               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0…
## $ peak_id_HIML               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_JANE               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_KANG               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_LANG               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_LHOT               <dbl> 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_LSHR               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0…
## $ peak_id_LSIS               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_MAKA               <dbl> 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_MANA               <dbl> 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_NANG               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_NPHU               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_NUPT               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0…
## $ peak_id_PUMO               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_RATC               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_id_TUKU               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_name_Cho.Oyu          <dbl> 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_name_Dhaulagiri.I     <dbl> 1, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_name_Everest          <dbl> 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1…
## $ peak_name_Manaslu          <dbl> 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ peak_name_other            <dbl> 0, 0, 1, 0, 0, 0, 0, 1, 1, 1, 1, 1, 0, 1, 0…
## $ season_Autumn              <dbl> 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0…
## $ season_Spring              <dbl> 1, 1, 0, 1, 0, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1…
## $ season_Winter              <dbl> 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ sex_F                      <dbl> 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0…
## $ sex_M                      <dbl> 1, 1, 1, 1, 1, 0, 1, 0, 1, 1, 1, 1, 1, 1, 1…
## $ citizenship_Australia      <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ citizenship_Japan          <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0…
## $ citizenship_Nepal          <dbl> 0, 0, 1, 0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0…
## $ citizenship_Spain          <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0…
## $ citizenship_UK             <dbl> 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ citizenship_USA            <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ citizenship_other          <dbl> 1, 1, 0, 0, 0, 1, 0, 1, 0, 1, 0, 0, 1, 0, 1…
## $ expedition_role_Climber    <dbl> 1, 0, 0, 1, 0, 1, 0, 0, 1, 1, 1, 1, 1, 0, 1…
## $ expedition_role_H.A.Worker <dbl> 0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0…
## $ expedition_role_Leader     <dbl> 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0…
## $ expedition_role_other      <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ hired_FALSE.               <dbl> 1, 1, 0, 1, 0, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1…
## $ hired_TRUE.                <dbl> 0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0…
## $ success_FALSE.             <dbl> 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0…
## $ success_TRUE.              <dbl> 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1…
## $ solo_FALSE.                <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1…
## $ solo_TRUE.                 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ oxygen_used_FALSE.         <dbl> 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1…
## $ oxygen_used_TRUE.          <dbl> 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…
## $ injured_FALSE.             <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1…
## $ injured_TRUE.              <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0…

Specify Model

library(usemodels) 

usemodels::use_xgboost(died ~., data = data_training)
## xgboost_recipe <- 
##   recipe(formula = died ~ ., data = data_training) %>% 
##   step_zv(all_predictors()) 
## 
## xgboost_spec <- 
##   boost_tree(trees = tune(), min_n = tune(), tree_depth = tune(), learn_rate = tune(), 
##     loss_reduction = tune(), sample_size = tune()) %>% 
##   set_mode("classification") %>% 
##   set_engine("xgboost") 
## 
## xgboost_workflow <- 
##   workflow() %>% 
##   add_recipe(xgboost_recipe) %>% 
##   add_model(xgboost_spec) 
## 
## set.seed(63393)
## xgboost_tune <-
##   tune_grid(xgboost_workflow, resamples = stop("add your rsample object"), grid = stop("add number of candidate points"))
xgboost_spec <- 
  boost_tree(trees = tune()) %>% 
  set_mode("classification") %>% 
  set_engine("xgboost") 

xgboost_workflow <- 
  workflow() %>% 
  add_recipe(xgboost_rec) %>% 
  add_model(xgboost_spec)

Tune Hyperparameters

doParallel::registerDoParallel()

set.seed(45034)
xgboost_tune <-
  tune_grid(xgboost_workflow, 
            resamples = data_cv, 
            grid = 5,
            control = control_grid(save_pred = TRUE))

Model Evaluation

Identifying optimal values for hyperparameters

collect_metrics(xgboost_tune)
## # A tibble: 10 × 7
##    trees .metric  .estimator  mean     n std_err .config             
##    <int> <chr>    <chr>      <dbl> <int>   <dbl> <chr>               
##  1   111 accuracy binary     0.646    10  0.0436 Preprocessor1_Model1
##  2   111 roc_auc  binary     0.606    10  0.0566 Preprocessor1_Model1
##  3   683 accuracy binary     0.646    10  0.0436 Preprocessor1_Model2
##  4   683 roc_auc  binary     0.598    10  0.0559 Preprocessor1_Model2
##  5  1015 accuracy binary     0.633    10  0.0460 Preprocessor1_Model3
##  6  1015 roc_auc  binary     0.610    10  0.0557 Preprocessor1_Model3
##  7  1205 accuracy binary     0.621    10  0.0479 Preprocessor1_Model4
##  8  1205 roc_auc  binary     0.617    10  0.0554 Preprocessor1_Model4
##  9  1840 accuracy binary     0.604    10  0.0490 Preprocessor1_Model5
## 10  1840 roc_auc  binary     0.618    10  0.0606 Preprocessor1_Model5
collect_predictions(xgboost_tune) %>%
    group_by(id) %>%
    roc_curve(died, .pred_died) %>%
    autoplot()

Fit the model for the last time

xgboost_last <- xgboost_workflow %>%
    finalize_workflow(select_best(xgboost_tune, metric = "accuracy")) %>%
    last_fit(data_split)
## → A | warning: There are new levels in a factor: MAKA91301, EVER96108, EVER10102, PUMO05103, TAWO15301, ANN199101, ANN191301, ANN380101, EVER12178, EVER95305, MANA86101, LHOT16102, EVER08105, EVER79301, CHOY11316, MAKA76101, ANNS88101, MAK285101, EVER98116, BARU11306, MANA12341, HIML12302, AMAD12347, CHOY08302, KANG18102, EVER19102, There are new levels in a factor: TAWO, ANNS, MAK2, BARU
## 
There were issues with some computations   A: x1

There were issues with some computations   A: x1
collect_metrics(xgboost_last)
## # A tibble: 2 × 4
##   .metric  .estimator .estimate .config             
##   <chr>    <chr>          <dbl> <chr>               
## 1 accuracy binary         0.538 Preprocessor1_Model1
## 2 roc_auc  binary         0.627 Preprocessor1_Model1
 collect_predictions(xgboost_last) %>%
     yardstick::conf_mat(died, .pred_class) %>%
     autoplot()

library(vip)
## 
## Attaching package: 'vip'
## The following object is masked from 'package:utils':
## 
##     vi
xgboost_last %>%
    workflows::extract_fit_engine() %>%
    vip()