Goal to predict attrition, employees who are likely to leave the company

library(tidyverse)
## ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.2 ──
## ✔ ggplot2 3.4.4     ✔ purrr   1.0.2
## ✔ tibble  3.2.1     ✔ dplyr   1.1.4
## ✔ tidyr   1.3.1     ✔ stringr 1.5.1
## ✔ readr   2.1.2     ✔ forcats 1.0.0
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag()    masks stats::lag()
library(parameters)
library(tidymodels) # for model building
## ── 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      1.0.9     
## ── Conflicts ───────────────────────────────────────── tidymodels_conflicts() ──
## ✖ scales::discard()  masks purrr::discard()
## ✖ dplyr::filter()    masks stats::filter()
## ✖ recipes::fixed()   masks stringr::fixed()
## ✖ dplyr::lag()       masks stats::lag()
## ✖ infer::p_value()   masks parameters::p_value()
## ✖ tune::parameters() masks dials::parameters(), parameters::parameters()
## ✖ yardstick::spec()  masks readr::spec()
## ✖ recipes::step()    masks stats::step()
## • Search for functions across packages at https://www.tidymodels.org/find/
library(textrecipes) # for preprocessing string
library(xgboost)
## 
## Attaching package: 'xgboost'
## 
## The following object is masked from 'package:dplyr':
## 
##     slice
library(tidytext)
library(Binarize)
## Loading required package: diptest
library(correlation)
library(correlationfunnel)
## ══ Using correlationfunnel? ════════════════════════════════════════════════════
## You might also be interested in applied data science training for business.
## </> Learn more at - www.business-science.io </>
library(doParallel)
## Loading required package: foreach
## 
## Attaching package: 'foreach'
## 
## The following objects are masked from 'package:purrr':
## 
##     accumulate, when
## 
## Loading required package: iterators
## Loading required package: parallel
library(LiblineaR)
library(ranger)
library(vip)
## 
## Attaching package: 'vip'
## 
## The following object is masked from 'package:utils':
## 
##     vi
library(themis)

members <- 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.

Clean Data

skimr::skim(members)
Data summary
Name members
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 ▁▁▂▇▇

Explore Data

data <- members %>%

 # Treat missing values 
  select(-death_cause, -injury_type, -highpoint_metres, -death_height_metres, -injury_height_metres) %>%
  na.omit() %>%
  
  # Log Transform Variables with pos-skewed Distribution
  mutate(across(where(is.logical), as.factor))
members %>%
  ggplot(aes(died)) +
  geom_bar()

Correlation Plot

# Step 1: Prepare data 
data_binarized_tbl <- data %>%
  select(-peak_name) %>%
  binarize()

data_binarized_tbl %>% glimpse()
## Rows: 72,985
## Columns: 71
## $ expedition_id__EVER88101       <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, …
## $ `member_id__KANG10101-01`      <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ `member_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__ANN4                  <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, …
## $ `year__-Inf_1992`              <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, …
## $ year__1992_2004                <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ year__2004_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, 1, 1, 1, 0, 0, 0, 0, 0, …
## $ season__Spring                 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 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, 0, 1, 0, 1, 0, 0, 1, 0, 1, …
## $ age__29_36                     <dbl> 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, …
## $ age__36_44                     <dbl> 1, 1, 0, 1, 0, 0, 1, 0, 0, 1, 0, 1, 0, …
## $ age__44_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, 1, 1, 1, 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__Netherlands       <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, 0, 0, 1, 0, 1, 1, 1, …
## $ citizenship__W_Germany         <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, …
## $ `citizenship__-OTHER`          <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ expedition_role__Climber       <dbl> 0, 0, 1, 0, 1, 1, 1, 1, 1, 1, 1, 0, 0, …
## $ expedition_role__Deputy_Leader <dbl> 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ expedition_role__Exp_Doctor    <dbl> 0, 0, 0, 1, 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> 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ `expedition_role__-OTHER`      <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 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, …
## $ success__FALSE                 <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, …
## $ success__TRUE                  <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, …
## $ 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__FALSE                    <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, …
## $ died__TRUE                     <dbl> 0, 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, …
## $ injured__TRUE                  <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
# Step 2: Correlate 
data_corr_tbl <- data_binarized_tbl %>%
  correlate(died__TRUE)
## Warning: correlate(): [Data Imbalance Detected] Consider sampling to balance the classes more than 5%
##   Column with imbalance: died__TRUE
data_corr_tbl
## # A tibble: 71 × 3
##    feature bin       correlation
##    <fct>   <chr>           <dbl>
##  1 died    FALSE         -1     
##  2 died    TRUE           1     
##  3 year    -Inf_1992      0.0519
##  4 peak_id ANN1           0.0336
##  5 success FALSE          0.0332
##  6 success TRUE          -0.0332
##  7 peak_id DHA1           0.0290
##  8 peak_id AMAD          -0.0281
##  9 peak_id CHOY          -0.0241
## 10 year    2004_2012     -0.0211
## # ℹ 61 more rows
# Step 3: Plot 
data_corr_tbl %>%
  plot_correlation_funnel()
## Warning: ggrepel: 41 unlabeled data points (too many overlaps). Consider
## increasing max.overlaps