Apply it to your data 8
Liam Smith
2026-04-03
Import data
library(tidyverse)## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr 1.1.4 ✔ readr 2.1.5
## ✔ forcats 1.0.0 ✔ stringr 1.5.2
## ✔ ggplot2 4.0.0 ✔ tibble 3.3.0
## ✔ lubridate 1.9.4 ✔ tidyr 1.3.1
## ✔ purrr 1.1.0
## ── 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 errorslibrary(correlationfunnel)## Warning: package 'correlationfunnel' was built under R version 4.5.2## ══ correlationfunnel Tip #1 ════════════════════════════════════════════════════
## Make sure your data is not overly imbalanced prior to using `correlate()`.
## If less than 5% imbalance, consider sampling. :)library(textrecipes)## Warning: package 'textrecipes' was built under R version 4.5.2## Loading required package: recipes
##
## Attaching package: 'recipes'
##
## The following object is masked from 'package:stringr':
##
## fixed
##
## The following object is masked from 'package:stats':
##
## stepdata <- readr::read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/main/data/2020/2020-11-03/ikea.csv')## New names:
## Rows: 3694 Columns: 14
## ── Column specification
## ──────────────────────────────────────────────────────── Delimiter: "," chr
## (7): name, category, old_price, link, other_colors, short_description, d... dbl
## (6): ...1, item_id, price, depth, height, width lgl (1): sellable_online
## 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.
## • `` -> `...1`skimr::skim(data)| Name | data |
| Number of rows | 3694 |
| Number of columns | 14 |
| _______________________ | |
| Column type frequency: | |
| character | 7 |
| logical | 1 |
| numeric | 6 |
| ________________________ | |
| Group variables | None |
Variable type: character
| skim_variable | n_missing | complete_rate | min | max | empty | n_unique | whitespace |
|---|---|---|---|---|---|---|---|
| name | 0 | 1 | 3 | 27 | 0 | 607 | 0 |
| category | 0 | 1 | 4 | 36 | 0 | 17 | 0 |
| old_price | 0 | 1 | 4 | 13 | 0 | 365 | 0 |
| link | 0 | 1 | 52 | 163 | 0 | 2962 | 0 |
| other_colors | 0 | 1 | 2 | 3 | 0 | 2 | 0 |
| short_description | 0 | 1 | 3 | 63 | 0 | 1706 | 0 |
| designer | 0 | 1 | 3 | 1261 | 0 | 381 | 0 |
Variable type: logical
| skim_variable | n_missing | complete_rate | mean | count |
|---|---|---|---|---|
| sellable_online | 0 | 1 | 0.99 | TRU: 3666, FAL: 28 |
Variable type: numeric
| skim_variable | n_missing | complete_rate | mean | sd | p0 | p25 | p50 | p75 | p100 | hist |
|---|---|---|---|---|---|---|---|---|---|---|
| …1 | 0 | 1.00 | 1846.50 | 1066.51 | 0 | 923.25 | 1846.5 | 2769.75 | 3693 | ▇▇▇▇▇ |
| item_id | 0 | 1.00 | 48632396.79 | 28887094.10 | 58487 | 20390574.00 | 49288078.0 | 70403572.75 | 99932615 | ▇▇▇▇▇ |
| price | 0 | 1.00 | 1078.21 | 1374.65 | 3 | 180.90 | 544.7 | 1429.50 | 9585 | ▇▁▁▁▁ |
| depth | 1463 | 0.60 | 54.38 | 29.96 | 1 | 38.00 | 47.0 | 60.00 | 257 | ▇▃▁▁▁ |
| height | 988 | 0.73 | 101.68 | 61.10 | 1 | 67.00 | 83.0 | 124.00 | 700 | ▇▂▁▁▁ |
| width | 589 | 0.84 | 104.47 | 71.13 | 1 | 60.00 | 80.0 | 140.00 | 420 | ▇▅▂▁▁ |
Clean data
data_clean <- data %>%
mutate(across(is.logical, as.factor),
sellable_online = factor(sellable_online, levels = c(FALSE, TRUE), labels = c("No", "Yes")),
sellable_online = fct_relevel(sellable_online, "Yes", "No")) %>%
select(-old_price, -link, -...1) %>%
na.omit() %>%
mutate(price = log(price))## Warning: There was 1 warning in `mutate()`.
## ℹ In argument: `across(is.logical, as.factor)`.
## Caused by warning:
## ! Use of bare predicate functions was deprecated in tidyselect 1.1.0.
## ℹ Please use wrap predicates in `where()` instead.
## # Was:
## data %>% select(is.logical)
##
## # Now:
## data %>% select(where(is.logical))Explore data
data_clean %>% count(sellable_online)## # A tibble: 2 × 2
## sellable_online n
## <fct> <int>
## 1 Yes 1886
## 2 No 13data_clean %>%
ggplot(aes(sellable_online)) +
geom_bar()
sellable_online vs price
data_clean %>%
ggplot(aes(sellable_online, price)) +
geom_boxplot()
Correlation plot
# Step 1: Binarize
data_binarized <- data_clean %>%
select(-item_id, -short_description) %>%
binarize()
data_binarized %>% glimpse()## Rows: 1,899
## Columns: 82
## $ name__ALGOT <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ name__BEKANT <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ name__BESTÅ <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `name__BILLY_/_OXBERG` <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ name__BRIMNES <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ name__BROR <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ name__EKET <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ name__GRÖNLID <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ name__HAVSTA <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ name__HAVSTEN <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ name__HEMNES <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ name__IVAR <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ name__JONAXEL <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ name__KALLAX <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ name__LIDHULT <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ name__LIXHULT <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ name__NORDLI <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ name__PAX <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ name__PLATSA <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `name__STUVA_/_FRITIDS` <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ name__TROFAST <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ name__VALLENTUNA <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ name__VIMLE <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `name__-OTHER` <dbl> 1, 1, 1, 1, 1, 1, 1, …
## $ category__Bar_furniture <dbl> 1, 1, 1, 1, 1, 1, 1, …
## $ category__Beds <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `category__Bookcases_&_shelving_units` <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `category__Cabinets_&_cupboards` <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ category__Chairs <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `category__Chests_of_drawers_&_drawer_units` <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `category__Children's_furniture` <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ category__Nursery_furniture <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ category__Outdoor_furniture <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `category__Sideboards,_buffets_&_console_tables` <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `category__Sofas_&_armchairs` <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `category__Tables_&_desks` <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `category__TV_&_media_furniture` <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ category__Wardrobes <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `category__-OTHER` <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `price__-Inf_5.68697535633982` <dbl> 1, 1, 0, 1, 1, 1, 0, …
## $ price__5.68697535633982_6.52209279817015 <dbl> 0, 0, 1, 0, 0, 0, 1, …
## $ price__6.52209279817015_7.37085996851068 <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ price__7.37085996851068_Inf <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ sellable_online__Yes <dbl> 1, 1, 1, 1, 1, 1, 1, …
## $ `sellable_online__-OTHER` <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ other_colors__No <dbl> 0, 1, 1, 1, 1, 1, 1, …
## $ other_colors__Yes <dbl> 1, 0, 0, 0, 0, 0, 0, …
## $ designer__Andreas_Fredriksson <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ designer__Carina_Bengs <dbl> 0, 0, 1, 0, 0, 0, 1, …
## $ designer__Carl_Öjerstam <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ designer__Ebba_Strandmark <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ designer__Ehlén_Johansson <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `designer__Ehlén_Johansson/IKEA_of_Sweden` <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ designer__Eva_Lilja_Löwenhielm <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ designer__Francis_Cayouette <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ designer__Gillis_Lundgren <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ designer__Henrik_Preutz <dbl> 1, 0, 0, 0, 0, 0, 0, …
## $ designer__IKEA_of_Sweden <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `designer__IKEA_of_Sweden/Ehlén_Johansson` <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `designer__IKEA_of_Sweden/Jon_Karlsson` <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ designer__Johan_Kroon <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ designer__Jon_Karlsson <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `designer__Jon_Karlsson/IKEA_of_Sweden` <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `designer__K_Hagberg/M_Hagberg` <dbl> 0, 0, 0, 1, 1, 1, 0, …
## $ designer__Mia_Lagerman <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ designer__Nike_Karlsson <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ designer__Ola_Wihlborg <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ designer__Studio_Copenhagen <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ designer__Tord_Björklund <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `designer__-OTHER` <dbl> 0, 1, 0, 0, 0, 0, 0, …
## $ `depth__-Inf_40` <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ depth__40_47 <dbl> 0, 0, 1, 1, 1, 1, 1, …
## $ depth__47_60 <dbl> 1, 1, 0, 0, 0, 0, 0, …
## $ depth__60_Inf <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `height__-Inf_71` <dbl> 0, 1, 0, 0, 0, 0, 0, …
## $ height__71_92 <dbl> 0, 0, 1, 0, 0, 0, 0, …
## $ height__92_171 <dbl> 1, 0, 0, 1, 1, 1, 1, …
## $ height__171_Inf <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ `width__-Inf_60` <dbl> 1, 0, 1, 1, 1, 1, 1, …
## $ width__60_93 <dbl> 0, 1, 0, 0, 0, 0, 0, …
## $ width__93_161.5 <dbl> 0, 0, 0, 0, 0, 0, 0, …
## $ width__161.5_Inf <dbl> 0, 0, 0, 0, 0, 0, 0, …# Step 2: Correlation
data_correlation <- data_binarized %>%
correlate(sellable_online__Yes)
data_correlation## # A tibble: 82 × 3
## feature bin correlation
## <fct> <chr> <dbl>
## 1 sellable_online Yes 1
## 2 sellable_online -OTHER -1
## 3 name TROFAST -0.332
## 4 category Children's_furniture -0.144
## 5 price -Inf_5.68697535633982 -0.143
## 6 category Nursery_furniture -0.128
## 7 width -Inf_60 -0.128
## 8 designer Studio_Copenhagen -0.112
## 9 depth -Inf_40 -0.0806
## 10 designer Francis_Cayouette -0.0573
## # ℹ 72 more rows# Step 3: Plot
data_correlation %>%
correlationfunnel::plot_correlation_funnel()## Warning: The `size` argument of `element_line()` is deprecated as of ggplot2 3.4.0.
## ℹ Please use the `linewidth` argument instead.
## ℹ The deprecated feature was likely used in the correlationfunnel package.
## Please report the issue at
## <https://github.com/business-science/correlationfunnel/issues>.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.## Warning: The `size` argument of `element_rect()` is deprecated as of ggplot2 3.4.0.
## ℹ Please use the `linewidth` argument instead.
## ℹ The deprecated feature was likely used in the correlationfunnel package.
## Please report the issue at
## <https://github.com/business-science/correlationfunnel/issues>.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.## Warning: ggrepel: 61 unlabeled data points (too many overlaps). Consider
## increasing max.overlaps
Model Building
Split data
library(tidymodels)## Warning: package 'tidymodels' was built under R version 4.5.2## ── Attaching packages ────────────────────────────────────── tidymodels 1.4.1 ──## ✔ broom 1.0.10 ✔ tailor 0.1.0
## ✔ dials 1.4.2 ✔ tune 2.0.1
## ✔ infer 1.1.0 ✔ workflows 1.3.0
## ✔ modeldata 1.5.1 ✔ workflowsets 1.1.1
## ✔ parsnip 1.4.1 ✔ yardstick 1.3.2
## ✔ rsample 1.3.1## Warning: package 'dials' was built under R version 4.5.2## Warning: package 'infer' was built under R version 4.5.2## Warning: package 'modeldata' was built under R version 4.5.2## Warning: package 'parsnip' was built under R version 4.5.2## Warning: package 'tailor' was built under R version 4.5.2## Warning: package 'tune' was built under R version 4.5.2## Warning: package 'workflows' was built under R version 4.5.2## Warning: package 'workflowsets' was built under R version 4.5.2## Warning: package 'yardstick' was built under R version 4.5.2## ── 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()set.seed(1234)
# data_clean <- data_clean %>% sample_n(100)
data_split <- initial_split(data_clean, strata = sellable_online)
data_train <- training(data_split)
data_test <- testing(data_split)
data_cv <- vfold_cv(data_train, v = 5, strata = sellable_online)
data_cv## # 5-fold cross-validation using stratification
## # A tibble: 5 × 2
## splits id
## <list> <chr>
## 1 <split [1139/285]> Fold1
## 2 <split [1139/285]> Fold2
## 3 <split [1139/285]> Fold3
## 4 <split [1139/285]> Fold4
## 5 <split [1140/284]> Fold5Preprocess data
library(themis)## Warning: package 'themis' was built under R version 4.5.2xgboost_rec <- recipes::recipe(sellable_online ~ ., data = data_train) %>%
update_role(item_id, new_role = "ID") %>%
step_tokenize(short_description) %>%
step_tokenfilter(short_description, max_tokens = 100) %>%
step_tf(short_description) %>%
step_other(name, designer, threshold = 0.02) %>%
step_dummy(all_nominal_predictors(), designer) %>%
step_zv(all_predictors()) %>%
step_smote(sellable_online, over_ratio = 1)
xgboost_rec %>% prep() %>% juice() %>% glimpse()## Rows: 2,828
## Columns: 145
## $ item_id <dbl> 70404875, 50406465, 9040…
## $ price <dbl> 4.859812, 4.859812, 5.00…
## $ depth <dbl> 44, 44, 44, 52, 51, 44, …
## $ height <dbl> 95, 95, 103, 114, 102, 1…
## $ width <dbl> 50, 50, 52, 43, 40, 52, …
## $ sellable_online <fct> Yes, Yes, Yes, Yes, Yes,…
## $ tf_short_description_1 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_147x147 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_150x44x236 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_150x60x236 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_150x66x236 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_2 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_200x60x236 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_200x66x236 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_25x51x70 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_3 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_35x35x35 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_4 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_41x61 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_5 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_50x51x70 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_6 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_60x50x128 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_63 <dbl> 1, 1, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_74 <dbl> 0, 0, 1, 0, 1, 1, 0, 0, …
## $ tf_short_description_75 <dbl> 0, 0, 0, 1, 0, 0, 0, 1, …
## $ tf_short_description_8 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_80x30x202 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_add <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_and <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_armchair <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_armrest <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_armrests <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_backrest <dbl> 1, 1, 1, 1, 1, 1, 0, 1, …
## $ tf_short_description_bar <dbl> 1, 1, 1, 1, 1, 1, 1, 1, …
## $ tf_short_description_baskets <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_bed <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_bedside <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_bench <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_bookcase <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_box <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_cabinet <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_cabinets <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_castors <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_chair <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_chaise <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_changing <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_chest <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ `tf_short_description_children's` <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_clothes <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_cm <dbl> 1, 1, 1, 1, 1, 1, 1, 1, …
## $ tf_short_description_combination <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_corner <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_cover <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_desk <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_door <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_doors <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_drawer <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_drawers <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_feet <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_foldable <dbl> 1, 1, 1, 0, 0, 1, 0, 0, …
## $ tf_short_description_for <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_frame <dbl> 0, 0, 0, 0, 1, 0, 0, 0, …
## $ tf_short_description_glass <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_high <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_highchair <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_in <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_inserts <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_junior <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_leg <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_legs <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_lock <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_longue <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_mesh <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_modular <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_mounted <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_of <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_on <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_outdoor <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_panel <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_plinth <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_rail <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_seat <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_section <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_sections <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_shelf <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_shelves <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_shelving <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_side <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_sliding <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_smart <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_sofa <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_stool <dbl> 1, 1, 1, 1, 1, 1, 1, 1, …
## $ tf_short_description_storage <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_table <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_top <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_tv <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_two <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_underframe <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_unit <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_upright <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_w <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_wall <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_wardrobe <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_wire <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tf_short_description_with <dbl> 1, 1, 1, 1, 1, 1, 0, 1, …
## $ name_EKET <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ name_GRÖNLID <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ name_IVAR <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ name_JONAXEL <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ name_LIDHULT <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ name_NORDLI <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ name_PAX <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ name_PLATSA <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ name_TROFAST <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ name_VIMLE <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ name_other <dbl> 1, 1, 1, 1, 1, 1, 1, 1, …
## $ category_Beds <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ category_Bookcases...shelving.units <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ category_Cabinets...cupboards <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ category_Café.furniture <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ category_Chairs <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ category_Chests.of.drawers...drawer.units <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ category_Children.s.furniture <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ category_Nursery.furniture <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ category_Outdoor.furniture <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ category_Room.dividers <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ category_Sideboards..buffets...console.tables <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ category_Sofas...armchairs <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ category_Tables...desks <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ category_Trolleys <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ category_TV...media.furniture <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ category_Wardrobes <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ other_colors_Yes <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ designer_Ehlén.Johansson <dbl> 0, 0, 0, 1, 0, 0, 0, 1, …
## $ designer_Ehlén.Johansson.IKEA.of.Sweden <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ designer_Francis.Cayouette <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ designer_Henrik.Preutz <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ designer_IKEA.of.Sweden <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ designer_IKEA.of.Sweden.Ehlén.Johansson <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ designer_Jon.Karlsson <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ designer_K.Hagberg.M.Hagberg <dbl> 1, 1, 1, 0, 0, 1, 0, 0, …
## $ designer_Ola.Wihlborg <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ designer_Studio.Copenhagen <dbl> 0, 0, 0, 0, 0, 0, 0, 0, …
## $ designer_other <dbl> 0, 0, 0, 0, 1, 0, 1, 0, …Specify model
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_rec) %>%
add_model(xgboost_spec) Tune hyperparameters
doParallel::registerDoParallel()
set.seed(1234)
xgboost_tune <-
tune_grid(xgboost_workflow,
resamples = data_cv,
grid = 5,
metrics = metric_set(roc_auc, accuracy),
control = control_grid(save_pred = TRUE))Model Evaluation
Identify optimal values for hyperparameters
collect_metrics(xgboost_tune)## # A tibble: 10 × 12
## trees min_n tree_depth learn_rate loss_reduction sample_size .metric
## <int> <int> <int> <dbl> <dbl> <dbl> <chr>
## 1 1 30 15 0.0750 0.0422 0.625 accuracy
## 2 1 30 15 0.0750 0.0422 0.625 roc_auc
## 3 500 21 1 0.316 0.0000562 1 accuracy
## 4 500 21 1 0.316 0.0000562 1 roc_auc
## 5 1000 2 8 0.0178 0.0000000001 0.5 accuracy
## 6 1000 2 8 0.0178 0.0000000001 0.5 roc_auc
## 7 1500 11 4 0.001 31.6 0.75 accuracy
## 8 1500 11 4 0.001 31.6 0.75 roc_auc
## 9 2000 40 11 0.00422 0.0000000750 0.875 accuracy
## 10 2000 40 11 0.00422 0.0000000750 0.875 roc_auc
## # ℹ 5 more variables: .estimator <chr>, mean <dbl>, n <int>, std_err <dbl>,
## # .config <chr>collect_predictions(xgboost_tune) %>%
group_by(id) %>%
roc_curve(sellable_online, .pred_Yes) %>%
autoplot()
Fit the model for the last time
xgboost_last <- xgboost_workflow %>%
finalize_workflow(select_best(xgboost_tune, metric = "accuracy")) %>%
last_fit(data_split)
collect_metrics(xgboost_last)## # A tibble: 3 × 4
## .metric .estimator .estimate .config
## <chr> <chr> <dbl> <chr>
## 1 accuracy binary 0.996 pre0_mod0_post0
## 2 roc_auc binary 0.968 pre0_mod0_post0
## 3 brier_class binary 0.00365 pre0_mod0_post0collect_predictions(xgboost_last) %>%
yardstick::conf_mat(sellable_online, .pred_class) %>%
autoplot()
Variable importance
library(vip)## Warning: package 'vip' was built under R version 4.5.3##
## Attaching package: 'vip'## The following object is masked from 'package:utils':
##
## vixgboost_last %>%
workflows::extract_fit_engine() %>%
vip()
Conclusion
The previous model had accuracy of 0.996 and AUC of 0.968.
- Feature transformation: normalized numeric data. No improvement.
- Feature transformation: YeoJohnson transformation. No improvement.
- Feature selection: PCA. Accuracy had a slight decrease 0.973 but AUC had a slight increase 0.971.
- Algorithm tuning: added mtry. No improvement.