Explore 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 1.9.2     ✔ 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
ratings <- read_csv("https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2022/2022-01-25/ratings.csv")
## Rows: 21831 Columns: 10
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: ","
## chr (3): name, url, thumbnail
## dbl (7): num, id, year, rank, average, bayes_average, users_rated
## 
## 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.
details <- read_csv("https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2022/2022-01-25/details.csv")
## Rows: 21631 Columns: 23
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: ","
## chr (10): primary, description, boardgamecategory, boardgamemechanic, boardg...
## dbl (13): num, id, yearpublished, minplayers, maxplayers, playingtime, minpl...
## 
## 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.
ratings_joined <-
    ratings %>%
    left_join(details, by = "id")
ggplot(ratings_joined, aes(average)) +
    geom_histogram(alpha = 0.8)
## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

ratings_joined %>%
    filter(!is.na(minage)) %>%
    mutate(minage = cut(minage, 4)) %>%
ggplot(aes(minage, average, fill = minage)) +
    geom_boxplot(alpha = 0.2, show.legend = FALSE)

Tune an xgboost model

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        1.0.4     ✔ workflows    1.1.3
## ✔ modeldata    1.2.0     ✔ workflowsets 1.0.1
## ✔ parsnip      1.1.1     ✔ yardstick    1.2.0
## ✔ 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 tidymodels_prefer() to resolve common conflicts.
set.seed(123)
game_split <-
  ratings_joined %>%
  select(name, average, matches("min|max"), boardgamecategory) %>%
  na.omit() %>%
  initial_split(strata = average)
game_train <- training(game_split)
game_test <- testing(game_split)

set.seed(234)
game_folds <- vfold_cv(game_train, strata = average)
game_folds
## #  10-fold cross-validation using stratification 
## # A tibble: 10 × 2
##    splits               id    
##    <list>               <chr> 
##  1 <split [14407/1602]> Fold01
##  2 <split [14407/1602]> Fold02
##  3 <split [14407/1602]> Fold03
##  4 <split [14408/1601]> Fold04
##  5 <split [14408/1601]> Fold05
##  6 <split [14408/1601]> Fold06
##  7 <split [14408/1601]> Fold07
##  8 <split [14408/1601]> Fold08
##  9 <split [14410/1599]> Fold09
## 10 <split [14410/1599]> Fold10
library(textrecipes)

split_category <- function(x) {
    x %>%
        str_split(", ") %>%
        map(str_remove_all, "[:punct:]") %>%
        map(str_squish) %>%
        map(str_to_lower) %>%
        map(str_replace_all, " ", "_")
}

game_rec <- 
    recipe(average ~ ., data = game_train) %>%
    update_role(name, new_role = "id") %>%
    step_tokenize(boardgamecategory, custom_token = split_category) %>%
    step_tokenfilter(boardgamecategory, max_tokens = 30) %>%
    step_tf(boardgamecategory)

game_prep <- prep(game_rec)
bake(game_prep, new_data = NULL)
## # A tibble: 16,009 × 37
##    name             minplayers maxplayers minplaytime maxplaytime minage average
##    <fct>                 <dbl>      <dbl>       <dbl>       <dbl>  <dbl>   <dbl>
##  1 Risk                      2          6         120         120     10    5.59
##  2 Monopoly                  2          8          60         180      8    4.37
##  3 UNO                       2         10          30          30      6    5.41
##  4 Apples to Apples          4         10          30          30     12    5.79
##  5 Zombies!!!                2          6          60          90     15    5.8 
##  6 Jenga                     1          8          20          20      6    5.62
##  7 The Game of Life          2          6          60          60      8    4.31
##  8 Battleship                2          2          30          30      8    4.66
##  9 Cranium                   4         16          60          60     13    5.68
## 10 Phase 10                  2          6          45          45      8    5.14
## # ℹ 15,999 more rows
## # ℹ 30 more variables: tf_boardgamecategory_abstract_strategy <int>,
## #   tf_boardgamecategory_action_dexterity <int>,
## #   tf_boardgamecategory_adventure <int>, tf_boardgamecategory_ancient <int>,
## #   tf_boardgamecategory_animals <int>, tf_boardgamecategory_bluffing <int>,
## #   tf_boardgamecategory_card_game <int>,
## #   tf_boardgamecategory_childrens_game <int>, …
xgb_spec <-
  boost_tree(
    trees = tune(),
    mtry = tune(),
    min_n = tune(),
    learn_rate = 0.01
  ) %>%
  set_engine("xgboost") %>%
  set_mode("regression")

xgb_wf <- workflow(game_rec, xgb_spec)
xgb_wf
## ══ Workflow ════════════════════════════════════════════════════════════════════
## Preprocessor: Recipe
## Model: boost_tree()
## 
## ── Preprocessor ────────────────────────────────────────────────────────────────
## 3 Recipe Steps
## 
## • step_tokenize()
## • step_tokenfilter()
## • step_tf()
## 
## ── Model ───────────────────────────────────────────────────────────────────────
## Boosted Tree Model Specification (regression)
## 
## Main Arguments:
##   mtry = tune()
##   trees = tune()
##   min_n = tune()
##   learn_rate = 0.01
## 
## Computational engine: xgboost
library(finetune)
doParallel::registerDoParallel()

set.seed(234)
xgb_game_rs <- 
tune_race_anova(
    xgb_wf, 
    game_folds,
    grid = 20, 
    control = control_race(verbose_elim = TRUE)
)
## i Creating pre-processing data to finalize unknown parameter: mtry
## ℹ Racing will minimize the rmse metric.
## ℹ Resamples are analyzed in a random order.
## ℹ Fold10: 6 eliminated; 14 candidates remain.
## 
## ℹ Fold06: 8 eliminated; 6 candidates remain.
## 
## ℹ Fold08: 2 eliminated; 4 candidates remain.
## 
## ℹ Fold01: 1 eliminated; 3 candidates remain.
## 
## ℹ Fold04: All but one parameter combination were eliminated.
xgb_game_rs
## # Tuning results
## # 10-fold cross-validation using stratification 
## # A tibble: 10 × 5
##    splits               id     .order .metrics          .notes          
##    <list>               <chr>   <int> <list>            <list>          
##  1 <split [14407/1602]> Fold03      1 <tibble [40 × 7]> <tibble [0 × 3]>
##  2 <split [14408/1601]> Fold05      2 <tibble [40 × 7]> <tibble [0 × 3]>
##  3 <split [14410/1599]> Fold10      3 <tibble [40 × 7]> <tibble [0 × 3]>
##  4 <split [14408/1601]> Fold06      4 <tibble [28 × 7]> <tibble [0 × 3]>
##  5 <split [14408/1601]> Fold08      5 <tibble [12 × 7]> <tibble [0 × 3]>
##  6 <split [14407/1602]> Fold01      6 <tibble [8 × 7]>  <tibble [0 × 3]>
##  7 <split [14408/1601]> Fold04      7 <tibble [6 × 7]>  <tibble [0 × 3]>
##  8 <split [14407/1602]> Fold02      8 <tibble [2 × 7]>  <tibble [0 × 3]>
##  9 <split [14408/1601]> Fold07     10 <tibble [2 × 7]>  <tibble [0 × 3]>
## 10 <split [14410/1599]> Fold09      9 <tibble [2 × 7]>  <tibble [0 × 3]>

Evaluation model

plot_race(xgb_game_rs)

show_best(xgb_game_rs)
## Warning: No value of `metric` was given; metric 'rmse' will be used.
## # A tibble: 1 × 9
##    mtry trees min_n .metric .estimator  mean     n std_err .config              
##   <int> <int> <int> <chr>   <chr>      <dbl> <int>   <dbl> <chr>                
## 1    14  1709    17 rmse    standard   0.735    10 0.00550 Preprocessor1_Model08
xgb_last <-
    xgb_wf %>% 
    finalize_workflow(select_best(xgb_game_rs, "rmse")) %>%
    last_fit(game_split)

xgb_last
## # Resampling results
## # Manual resampling 
## # A tibble: 1 × 6
##   splits               id              .metrics .notes   .predictions .workflow 
##   <list>               <chr>           <list>   <list>   <list>       <list>    
## 1 <split [16009/5339]> train/test spl… <tibble> <tibble> <tibble>     <workflow>
library(vip)
## 
## Attaching package: 'vip'
## The following object is masked from 'package:utils':
## 
##     vi
xgb_fit <- extract_workflow(xgb_last)
vip(xgb_fit, geom = "point", num_features = 12)

library(SHAPforxgboost)

game_shap <-
  shap.prep(
    xgb_model = extract_fit_engine(xgb_fit),
    X_train = bake(game_prep,
      has_role("predictor"),
      new_data = NULL,
      composition = "matrix"
    )
  )

shap.plot.summary(game_shap)

shap.plot.dependence(
    game_shap,
    x = "minage",
    color_feature = "minplayers",
    size0 = 1.2,
    smooth = FALSE, add_hist = TRUE
)

Questions

  1. Question and Data:

    • What is the research question? Clearly state the research question you aim to address using the new dataset. Can we predict ratings for board games based on the main characteristics like number of players and game category? How are the ratings distributed?

    • Describe the data briefly: Provide an overview of the new data set, highlighting its key characteristics and dimensions. The data sets are all about board games and board game ratings. Various information about board games are included in the data sets. The dimensions of the ratings.csv data set are 21,831 rows and 10 columns. The dimensions of the details.csv data set is 21,631 rows and 23 columns. The key variables are “name” which describes the name of the board games. “average” is the average rating of the board game. “year” is the year the board game was released. “boardgamecategory” is a description of the type of category that the boardgame falls into. “minplayers” and “maxplayers” are also key variables that provide insight on how many minimum players and maximum players are needed to play each game.

    • What are the characteristics of the key variables used in the analysis? Describe the primary variables of interest in the dataset and their characteristics. “average” is numerical data and was used as the target variable in the XGBoost model. “name” is character data and is important because the name of the board game is essential for referencing the games. “year” is numeric data and could be used to look at trends over time such as if old games rank better or worse than new games. “boardgamecategory” is character data and was used to determine if the board game category had any connection to the rating of the board game. “minplayers” and “maxplayers” are numerical data, and these variables tell us how inclusive or exclusive a game is which could influence ratings.

  2. Data Exploration and Transformation:
    • Describe the differences between the original data and the data transformed for modeling. Why? Explain any preprocessing or transformations performed on the new dataset compared to the original data. Discuss why these changes were necessary or beneficial. The original data set contains raw information about board games including id, average rating, board game name, the year it was released, the category, the minimum and maximum players needed to play, and the minimum age for players. The data set is split between two data frames called details and ratings. The new data has been transformed. The ratings and details datasets are mereged using the “id” column so that both the average rating and the board game details can be seen. This is necessary because the model could not be trained to predict the rating based off the details if it could not see both at the same time. In the code along we tokenized “boardgamecetgory” so that each category became a column of its own and each new column would indicate whether the board game belongs to that category. This was needed because XGBoost requires numeric data, so tokenizing transformed textual data into a format that the model can use. The “name” and “id” columns were removed from the new data because they are not numerical data, and are really more of an identifier than a predictor. These categories did not contribute to how the model precits and leaving them made the model more complicated than necessary. The “average” column was separated from the predictors in the new data so that the model could be trained to predict the average based on the features. This was needed so that the model new specifically what we were trying to predict based off of what key variables.
  3. Data Preparation and Modeling:

    • What are the names of data preparation steps mentioned in the video? List and describe any data preparation steps or techniques mentioned in the CA video that you applied to the new dataset. A preparation step used on the code is na.omit() which we used to remove rows with missing values. We also used as.integer to convert the “year” column to integer data. This way the data will be treated as numeric not character. We also filtered the data to only include rows where the year was greater than or equal to 2000 using the subset function. This way we were using more recent records. We also used (x-min(x))/(max(x)-min(x)) to bring the values between 0 and 1. Value_squared() and value_log() were used based on the “value” column. The squared coulmn captures non-linear relationships and the log-transformed column can look at skewness in the data. Lastly, we used head() to select only the first 1,000 rows of the data set which made the data set more manageable.

    • What is the name of the machine learning model(s) used in the analysis? Specify the machine learning model(s) you employed for your analysis and briefly explain their relevance to the research question. The machine learning model used in the analysis is XGBoost. We used set_engine(xgboost) with set_mode(regression) to make a regression model that predicts board game ratings.

  4. Model Evaluation:
    • What metrics are used in the model evaluation? Detail the evaluation metrics you used to assess the performance of your machine learning model(s) on the new dataset. Discuss the significance of these metrics in the context of your research question. The metrics we used in the model evaluation are Root mean square error (RMSE). We used this in the code select_best(xgb_game_rs, “rmse”). RMSE measures the average error between the predicted board game rating compared to the actual board game rating. The lower the RMSE, the better the regression model is at predicting the average.
  5. Conclusion:
    • What are the major findings? Summarize the key findings and insights obtained from your analysis of the new dataset. XGBoost performs well in predicting the board game ratings because the RMSE score was on the lower end which is good. The variables “minage” and “boardgamecategory” could be good predictors for predicitng the board game ratings. This model could be helpful for people to use to decide which board game they want to buy at the store or pick out from the family board game collection. It could also help people who make board games figure out which aspects of board games seem to give board game’s higher rating amongst players. The SHAP values may show board game developers which parts of board games seem to give them higher ratings, such as min players, max players, board game category, etc. Using these to predict the average could then show board game developers that maybe having the minimum players at 3 and maximum players at 8 gives the board game the best ratings, thus making it a possible best seller.