Apply 4
Bella Kalinyak
2024-10-03
Import Data
ikea <- readr::read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/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(ikea)| Name | ikea |
| 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 | ▇▅▂▁▁ |
data <- ikea %>%
mutate(across(is.logical, as.factor)) %>%
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
category
data %>%
ggplot(aes(category, price)) +
geom_point()
other colors
data %>%
ggplot(aes(y = price, x = other_colors)) +
geom_point()
product height
data %>%
ggplot(aes(price, height)) +
geom_point()
product depth
data %>%
ggplot(aes(price, depth)) +
geom_point()
product width
data %>%
ggplot(aes(price, width)) +
geom_point()
title
# Step 1: Prepare data
data_binarized_tbl <- data %>%
select(-item_id, -short_description,) %>%
binarize
data_binarized_tbl %>% 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__TRUE <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: Correlate
data_corr_tbl <- data_binarized_tbl %>%
correlate(price__7.37085996851068_Inf)
data_corr_tbl## # A tibble: 82 × 3
## feature bin correlation
## <fct> <chr> <dbl>
## 1 price 7.37085996851068_Inf 1
## 2 width 161.5_Inf 0.579
## 3 depth 60_Inf 0.447
## 4 category Sofas_&_armchairs 0.379
## 5 width -Inf_60 -0.374
## 6 price -Inf_5.68697535633982 -0.336
## 7 price 6.52209279817015_7.37085996851068 -0.333
## 8 price 5.68697535633982_6.52209279817015 -0.331
## 9 name PAX 0.302
## 10 category Wardrobes 0.279
## # ℹ 72 more rows#Step 3: Plot
data_corr_tbl %>%
plot_correlation_funnel()## Warning: ggrepel: 58 unlabeled data points (too many overlaps). Consider
## increasing max.overlaps
Build models
# data <- sample_n(data, 100)
# Split into train and test dataset
set.seed(1234)
data_split <- rsample::initial_split(data)
data_train <- training(data_split)
data_test <- testing(data_split)
# Further split training dataset for cross-validation
set.seed(2345)
data_cv <- rsample::vfold_cv(data_train)
data_cv## # 10-fold cross-validation
## # A tibble: 10 × 2
## splits id
## <list> <chr>
## 1 <split [1281/143]> Fold01
## 2 <split [1281/143]> Fold02
## 3 <split [1281/143]> Fold03
## 4 <split [1281/143]> Fold04
## 5 <split [1282/142]> Fold05
## 6 <split [1282/142]> Fold06
## 7 <split [1282/142]> Fold07
## 8 <split [1282/142]> Fold08
## 9 <split [1282/142]> Fold09
## 10 <split [1282/142]> Fold10library(usemodels)
usemodels::use_xgboost(price ~ ., data = data_train)## xgboost_recipe <-
## recipe(formula = price ~ ., data = data_train) %>%
## 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(41506)
## xgboost_tune <-
## tune_grid(xgboost_workflow, resamples = stop("add your rsample object"), grid = stop("add number of candidate points"))# Specify recipe
xgboost_recipe <-
recipe(formula = price ~ ., data = data_train) %>%
recipes::update_role(item_id, new_role = "id variable") %>%
step_tokenize(short_description) %>%
step_tokenfilter(short_description, max_tokens = 100) %>%
step_tfidf(short_description) %>%
step_other(designer, name, category) %>%
step_dummy(designer, name, category, other_colors, sellable_online, one_hot = TRUE) %>%
step_YeoJohnson(height, width, depth) %>%
step_normalize(all_numeric_predictors())
xgboost_recipe %>%
prep() %>% juice() %>% glimpse()## Rows: 1,424
## Columns: 124
## $ item_id <dbl> 19282962, 29320911, 49276549…
## $ depth <dbl> 2.52105533, 0.37916256, -0.1…
## $ height <dbl> -0.35419124, 1.68526717, -0.…
## $ width <dbl> 1.00466750, 0.01405564, -1.2…
## $ price <dbl> 7.595890, 7.833996, 6.429719…
## $ tfidf_short_description_1 <dbl> -0.1104286, -0.1104286, -0.1…
## $ tfidf_short_description_10 <dbl> -0.0647721, -0.0647721, -0.0…
## $ tfidf_short_description_120x40x64 <dbl> -0.06436942, -0.06436942, -0…
## $ tfidf_short_description_140x200 <dbl> -0.06483335, -0.06483335, -0…
## $ tfidf_short_description_147x147 <dbl> -0.07789055, -0.07789055, -0…
## $ tfidf_short_description_150x44x236 <dbl> -0.07463946, -0.07463946, -0…
## $ tfidf_short_description_150x60x236 <dbl> -0.07906935, -0.07906935, -0…
## $ tfidf_short_description_150x66x236 <dbl> -0.1187935, -0.1187935, -0.1…
## $ tfidf_short_description_2 <dbl> 5.0536418, -0.2738151, 2.922…
## $ tfidf_short_description_200x66x236 <dbl> -0.09595235, -0.09595235, -0…
## $ tfidf_short_description_25x51x70 <dbl> -0.06694652, -0.06694652, -0…
## $ tfidf_short_description_3 <dbl> -0.2590367, -0.2590367, -0.2…
## $ tfidf_short_description_4 <dbl> -0.1826156, -0.1826156, -0.1…
## $ tfidf_short_description_41x101 <dbl> -0.07491564, -0.07491564, -0…
## $ tfidf_short_description_41x61 <dbl> -0.07492126, -0.07492126, -0…
## $ tfidf_short_description_5 <dbl> -0.1120137, -0.1120137, -0.1…
## $ tfidf_short_description_50x51x70 <dbl> -0.06655607, -0.06655607, -0…
## $ tfidf_short_description_6 <dbl> -0.1166792, -0.1166792, -0.1…
## $ tfidf_short_description_60x50x128 <dbl> -0.06762634, -0.06762634, -0…
## $ tfidf_short_description_61x101 <dbl> -0.07006119, -0.07006119, -0…
## $ tfidf_short_description_74 <dbl> -0.09920561, -0.09920561, -0…
## $ tfidf_short_description_75 <dbl> -0.07026055, -0.07026055, -0…
## $ tfidf_short_description_8 <dbl> -0.07424754, -0.07424754, -0…
## $ tfidf_short_description_80x30x202 <dbl> -0.06624681, -0.06624681, -0…
## $ tfidf_short_description_99x44x56 <dbl> -0.06960254, -0.06960254, -0…
## $ tfidf_short_description_and <dbl> -0.08627569, -0.08627569, -0…
## $ tfidf_short_description_armchair <dbl> -0.1618879, -0.1618879, -0.1…
## $ tfidf_short_description_armrest <dbl> -0.07145785, -0.07145785, -0…
## $ tfidf_short_description_armrests <dbl> -0.1158378, -0.1158378, -0.1…
## $ tfidf_short_description_backrest <dbl> -0.1428615, -0.1428615, -0.1…
## $ tfidf_short_description_bar <dbl> -0.1454605, -0.1454605, -0.1…
## $ tfidf_short_description_baskets <dbl> -0.1425805, -0.1425805, -0.1…
## $ tfidf_short_description_bed <dbl> -0.2448542, -0.2448542, -0.2…
## $ tfidf_short_description_bench <dbl> -0.1929018, -0.1929018, -0.1…
## $ tfidf_short_description_bookcase <dbl> -0.1917741, -0.1917741, -0.1…
## $ tfidf_short_description_box <dbl> -0.07323435, -0.07323435, -0…
## $ tfidf_short_description_cabinet <dbl> -0.2548607, -0.2548607, -0.2…
## $ tfidf_short_description_cabinets <dbl> -0.06996899, -0.06996899, -0…
## $ tfidf_short_description_castors <dbl> -0.1183604, -0.1183604, -0.1…
## $ tfidf_short_description_chair <dbl> -0.2641368, -0.2641368, -0.2…
## $ tfidf_short_description_chaise <dbl> -0.09975977, -0.09975977, -0…
## $ tfidf_short_description_changing <dbl> -0.08762224, -0.08762224, -0…
## $ tfidf_short_description_chest <dbl> -0.2427739, -0.2427739, 4.03…
## $ `tfidf_short_description_children's` <dbl> -0.1109746, -0.1109746, -0.1…
## $ tfidf_short_description_clothes <dbl> -0.08137938, -0.08137938, -0…
## $ tfidf_short_description_cm <dbl> -1.19077286, 0.67884200, -0.…
## $ tfidf_short_description_combination <dbl> -0.3923715, 3.2920136, -0.39…
## $ tfidf_short_description_corner <dbl> -0.1735898, -0.1735898, -0.1…
## $ tfidf_short_description_cover <dbl> -0.0874799, -0.0874799, -0.0…
## $ tfidf_short_description_desk <dbl> -0.1262777, -0.1262777, -0.1…
## $ tfidf_short_description_door <dbl> -0.1236528, -0.1236528, -0.1…
## $ tfidf_short_description_doors <dbl> -0.3065925, -0.3065925, -0.3…
## $ tfidf_short_description_drawer <dbl> -0.1140866, -0.1140866, -0.1…
## $ tfidf_short_description_drawers <dbl> -0.3098302, -0.3098302, 3.32…
## $ tfidf_short_description_feet <dbl> -0.07972426, -0.07972426, -0…
## $ tfidf_short_description_foldable <dbl> -0.07006119, -0.07006119, -0…
## $ tfidf_short_description_for <dbl> -0.1016577, -0.1016577, -0.1…
## $ tfidf_short_description_frame <dbl> -0.2404379, -0.2404379, -0.2…
## $ tfidf_short_description_glass <dbl> -0.2229602, -0.2229602, -0.2…
## $ tfidf_short_description_highchair <dbl> -0.09382492, -0.09382492, -0…
## $ tfidf_short_description_in <dbl> -0.1139613, -0.1139613, -0.1…
## $ tfidf_short_description_inserts <dbl> -0.09931303, -0.09931303, -0…
## $ tfidf_short_description_junior <dbl> -0.1069594, -0.1069594, -0.1…
## $ tfidf_short_description_leg <dbl> -0.07995361, -0.07995361, -0…
## $ tfidf_short_description_legs <dbl> -0.08587662, -0.08587662, -0…
## $ tfidf_short_description_lock <dbl> -0.102278, -0.102278, -0.102…
## $ tfidf_short_description_longue <dbl> -0.09975977, -0.09975977, -0…
## $ tfidf_short_description_mesh <dbl> -0.1045224, -0.1045224, -0.1…
## $ tfidf_short_description_modular <dbl> -0.1372447, -0.1372447, -0.1…
## $ tfidf_short_description_module <dbl> -0.06960254, -0.06960254, -0…
## $ tfidf_short_description_mounted <dbl> -0.1380174, -0.1380174, -0.1…
## $ tfidf_short_description_of <dbl> -0.2444932, -0.2444932, 3.62…
## $ tfidf_short_description_on <dbl> -0.1219007, -0.1219007, -0.1…
## $ tfidf_short_description_outdoor <dbl> -0.1938584, -0.1938584, -0.1…
## $ tfidf_short_description_panel <dbl> -0.07016879, -0.07016879, -0…
## $ tfidf_short_description_plinth <dbl> -0.08108993, -0.08108993, -0…
## $ tfidf_short_description_rail <dbl> -0.06910249, -0.06910249, -0…
## $ tfidf_short_description_seat <dbl> 3.3170326, -0.3443572, -0.34…
## $ tfidf_short_description_section <dbl> -0.1761292, -0.1761292, -0.1…
## $ tfidf_short_description_sections <dbl> -0.1292483, -0.1292483, -0.1…
## $ tfidf_short_description_shelf <dbl> -0.1053796, -0.1053796, -0.1…
## $ tfidf_short_description_shelves <dbl> -0.1662117, -0.1662117, -0.1…
## $ tfidf_short_description_shelving <dbl> -0.2303551, -0.2303551, -0.2…
## $ tfidf_short_description_sliding <dbl> -0.07498137, -0.07498137, -0…
## $ tfidf_short_description_smart <dbl> -0.102278, -0.102278, -0.102…
## $ tfidf_short_description_sofa <dbl> 3.344695, -0.338867, -0.3388…
## $ tfidf_short_description_step <dbl> -0.07248683, -0.07248683, -0…
## $ tfidf_short_description_stool <dbl> -0.1598313, -0.1598313, -0.1…
## $ tfidf_short_description_storage <dbl> -0.3917943, -0.3917943, -0.3…
## $ tfidf_short_description_table <dbl> -0.139628, -0.139628, -0.139…
## $ tfidf_short_description_top <dbl> -0.07246422, -0.07246422, -0…
## $ tfidf_short_description_tv <dbl> -0.2361125, -0.2361125, -0.2…
## $ tfidf_short_description_underframe <dbl> -0.06172056, -0.06172056, -0…
## $ tfidf_short_description_unit <dbl> -0.2955415, -0.2955415, -0.2…
## $ tfidf_short_description_upright <dbl> -0.1161283, -0.1161283, -0.1…
## $ tfidf_short_description_w <dbl> -0.1942499, -0.1942499, -0.1…
## $ tfidf_short_description_wall <dbl> -0.2052322, -0.2052322, -0.2…
## $ tfidf_short_description_wardrobe <dbl> -0.3291643, 2.3781334, -0.32…
## $ tfidf_short_description_wire <dbl> -0.09063481, -0.09063481, -0…
## $ tfidf_short_description_with <dbl> -0.4838599, -0.4838599, -0.4…
## $ designer_IKEA.of.Sweden <dbl> -0.5544417, -0.5544417, -0.5…
## $ designer_Ola.Wihlborg <dbl> 4.3318115, -0.2306882, 4.331…
## $ designer_other <dbl> -1.5801952, 0.6323888, -1.58…
## $ name_BESTÅ <dbl> -0.2747283, -0.2747283, -0.2…
## $ name_PAX <dbl> -0.2455003, -0.2455003, -0.2…
## $ name_other <dbl> 0.3814619, 0.3814619, 0.3814…
## $ category_Bookcases...shelving.units <dbl> -0.4901523, -0.4901523, -0.4…
## $ category_Cabinets...cupboards <dbl> -0.3518917, -0.3518917, -0.3…
## $ category_Chairs <dbl> -0.3656009, -0.3656009, -0.3…
## $ category_Chests.of.drawers...drawer.units <dbl> -0.2672679, -0.2672679, 3.73…
## $ category_Sofas...armchairs <dbl> 2.881207, -0.346833, -0.3468…
## $ category_Tables...desks <dbl> -0.2323706, -0.2323706, -0.2…
## $ category_TV...media.furniture <dbl> -0.2422701, -0.2422701, -0.2…
## $ category_Wardrobes <dbl> -0.3417319, 2.9242158, -0.34…
## $ category_other <dbl> -0.4879483, -0.4879483, -0.4…
## $ other_colors_No <dbl> 0.9570323, 0.9570323, -1.044…
## $ other_colors_Yes <dbl> -0.9570323, -0.9570323, 1.04…
## $ sellable_online_FALSE. <dbl> -0.09215539, -0.09215539, -0…
## $ sellable_online_TRUE. <dbl> 0.09215539, 0.09215539, 0.09…#Specify model
xgboost_spec <-
boost_tree(trees = tune(), min_n = tune(), learn_rate = tune()) %>%
set_mode("regression") %>%
set_engine("xgboost")
# Combine recipe and model using workflow
xgboost_workflow <-
workflow() %>%
add_recipe(xgboost_recipe) %>%
add_model(xgboost_spec)
# Tune hyperparameters
set.seed(91707)
xgboost_tune <-
tune_grid(xgboost_workflow, resamples = data_cv, grid = 5)Evaluate Models
tune::show_best(xgboost_tune, metric = "rmse")## # A tibble: 5 × 9
## trees min_n learn_rate .metric .estimator mean n std_err .config
## <int> <int> <dbl> <chr> <chr> <dbl> <int> <dbl> <chr>
## 1 1191 17 0.0746 rmse standard 0.386 10 0.0156 Preprocessor1_M…
## 2 558 37 0.133 rmse standard 0.400 10 0.0116 Preprocessor1_M…
## 3 1799 17 0.0134 rmse standard 0.400 10 0.0134 Preprocessor1_M…
## 4 1436 30 0.00381 rmse standard 0.463 10 0.0133 Preprocessor1_M…
## 5 146 9 0.00173 rmse standard 4.73 10 0.0287 Preprocessor1_M…# Update the model by selecting the best hyperparameter
xgboost_fw <- tune::finalize_workflow(xgboost_workflow,
tune::select_best(xgboost_tune, metric = "rmse"))
# Fit the model of the entire training data and test it on the test data.
data_fit <- tune::last_fit(xgboost_fw, data_split)tune::collect_metrics(data_fit)## # A tibble: 2 × 4
## .metric .estimator .estimate .config
## <chr> <chr> <dbl> <chr>
## 1 rmse standard 0.343 Preprocessor1_Model1
## 2 rsq standard 0.924 Preprocessor1_Model1tune::collect_metrics(data_fit)## # A tibble: 2 × 4
## .metric .estimator .estimate .config
## <chr> <chr> <dbl> <chr>
## 1 rmse standard 0.343 Preprocessor1_Model1
## 2 rsq standard 0.924 Preprocessor1_Model1tune::collect_predictions(data_fit) %>%
ggplot(aes(price, .pred)) +
geom_point(alpha = 0.3, fill = "midnightblue") +
geom_abline(lty = 2, color = "gray50") +
coord_fixed()
Make Predictions
The dots are mostly centered around the prediction line. This model would make good predictions.
Changes
I added “one_hot = TRUE” to the step_dummy function, and “category” to the step_other function. I also added step_normalize to the recipe. The xgboost model showed to be the best with an RMSE value of 0.343 and an R squared value of 0.924. The rand_forest model had an RMSE value of 0.416 and an R sqaured value of 0.898. The svm_linear model had an RMSE value of 0.642 and an R sqaured value of 0.730.