Goal: to predict the rental market prices in the SF rental market Click here for the data.

Import Data

rent <- readr::read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2022/2022-07-05/rent.csv')

## Rows: 200796 Columns: 17
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: ","
## chr (8): post_id, nhood, city, county, address, title, descr, details
## dbl (9): date, year, price, beds, baths, sqft, room_in_apt, lat, lon
## 
## 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.

skimr::skim(rent)

Data summary
Name	rent
Number of rows	200796
Number of columns	17
_______________________
Column type frequency:
character	8
numeric	9
________________________
Group variables	None

Variable type: character

skim_variable	n_missing	complete_rate	min	max	n_unique
post_id	0	1.00	9	14	200796
nhood	0	1.00	4	43	167
city	0	1.00	5	19	104
county	1394	0.99	4	13	10
address	196888	0.02	1	38	2869
title	2517	0.99	2	298	184961
descr	197542	0.02	13	16975	3025
details	192780	0.04	4	595	7667

Variable type: numeric

skim_variable	n_missing	complete_rate	mean	sd	p0	p25	p50	p75	p100	hist
date	0	1.00	20095718.38	44694.07	20000902.00	20050227.00	20110924.00	20120805.0	20180717.00	▁▇▁▆▃
year	0	1.00	2009.51	4.48	2000.00	2005.00	2011.00	2012.0	2018.00	▁▇▁▆▃
price	0	1.00	2135.36	1427.75	220.00	1295.00	1800.00	2505.0	40000.00	▇▁▁▁▁
beds	6608	0.97	1.89	1.08	0.00	1.00	2.00	3.0	12.00	▇▂▁▁▁
baths	158121	0.21	1.68	0.69	1.00	1.00	2.00	2.0	8.00	▇▁▁▁▁
sqft	136117	0.32	1201.83	5000.22	80.00	750.00	1000.00	1360.0	900000.00	▇▁▁▁▁
room_in_apt	0	1.00	0.00	0.04	0.00	0.00	0.00	0.0	1.00	▇▁▁▁▁
lat	193145	0.04	37.67	0.35	33.57	37.40	37.76	37.8	40.43	▁▁▅▇▁
lon	196484	0.02	-122.21	0.78	-123.20	-122.42	-122.26	-122.0	-74.20	▇▁▁▁▁

data <- rent %>%
    
    # Treat missing values
    select(-address, -descr, -details, -lat, -lon, -date, -year, -room_in_apt) %>%
    na.omit() %>%
    
    # log transform variables with pos-skewed distribution
    mutate(price = log(price))

Explore Data

Identify good predictors.

sqft

data %>%
    ggplot(aes(price, sqft)) +
    scale_y_log10() +
    geom_point()

beds

data %>%
    ggplot(aes(price, as.factor(beds))) +
    geom_boxplot()

title

data %>%
    
    # tokenize title
    unnest_tokens(output = word, input = title) %>%
    
    # calculate average rent per word
    group_by(word) %>%
    summarise(price = mean(price),
              n     = n()) %>%
    ungroup() %>%
    
    filter(n > 10, !str_detect(word, "\\d")) %>%
    slice_max(order_by = price, n =20) %>%
    
    # Plot
    ggplot(aes(price, fct_reorder(word, price))) +
    geom_point() +
    
    labs(y = "Words in Title")

EDA shortcut

# Step 1: Prepare data
data_binarized_tbl <- data %>%
    select(-post_id, -title) %>%
    binarize()

data_binarized_tbl %>% glimpse()

## Rows: 14,394
## Columns: 85
## $ nhood__campbell                             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ `nhood__concord_/_pleasant_hill_/_martinez` <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__cupertino                            <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__daly_city                            <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ `nhood__danville_/_san_ramon`               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ `nhood__dublin_/_pleasanton`                <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ `nhood__fairfield_/_vacaville`              <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__foster_city                          <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ `nhood__hayward_/_castro_valley`            <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__milpitas                             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__mountain_view                        <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__napa_county                          <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__palo_alto                            <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__petaluma                             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ `nhood__pittsburg_/_antioch`                <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ `nhood__rohnert_pk_/_cotati`                <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__san_francisco                        <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__san_jose_central                     <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__san_jose_east                        <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__san_jose_north                       <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__san_jose_south                       <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__san_jose_west                        <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__san_mateo                            <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__san_rafael                           <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__santa_clara                          <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__santa_cruz                           <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__santa_rosa                           <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ `nhood__SOMA_/_south_beach`                 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__sunnyvale                            <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ nhood__union_city                           <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ `nhood__vallejo_/_benicia`                  <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ `nhood__willow_glen_/_cambrian`             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ `nhood__-OTHER`                             <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1,…
## $ city__cambrian                              <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__campbell                              <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__concord                               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__cupertino                             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__daly_city                             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__dublin                                <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__fairfield                             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__foster_city                           <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__hayward                               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__milpitas                              <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__mountain_view                         <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__napa_county                           <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__oakland                               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__palo_alto                             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__petaluma                              <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__pittsburg                             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__rohnert_park                          <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__san_francisco                         <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__san_jose                              <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__san_mateo                             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__san_rafael                            <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__san_ramon                             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__santa_clara                           <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__santa_cruz                            <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__santa_rosa                            <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__sunnyvale                             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__union_city                            <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ city__vallejo                               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ `city__-OTHER`                              <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1,…
## $ county__alameda                             <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1,…
## $ county__contra_costa                        <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ county__marin                               <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ county__napa                                <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ county__san_francisco                       <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ county__san_mateo                           <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ county__santa_clara                         <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ county__santa_cruz                          <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ county__solano                              <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ county__sonoma                              <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ `price__-Inf_7.52294091807237`              <dbl> 0, 1, 0, 1, 0, 1, 1, 0, 0,…
## $ price__7.52294091807237_7.80384330353877    <dbl> 0, 0, 1, 0, 0, 0, 0, 1, 1,…
## $ price__7.80384330353877_8.07868822922987    <dbl> 1, 0, 0, 0, 1, 0, 0, 0, 0,…
## $ price__8.07868822922987_Inf                 <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ `beds__-Inf_2`                              <dbl> 0, 1, 0, 1, 1, 1, 0, 0, 1,…
## $ beds__2_3                                   <dbl> 0, 0, 1, 0, 0, 0, 1, 1, 0,…
## $ beds__3_Inf                                 <dbl> 1, 0, 0, 0, 0, 0, 0, 0, 0,…
## $ `baths__-Inf_2`                             <dbl> 0, 1, 1, 1, 1, 1, 0, 0, 1,…
## $ baths__2_Inf                                <dbl> 1, 0, 0, 0, 0, 0, 1, 1, 0,…
## $ `sqft__-Inf_887`                            <dbl> 0, 1, 0, 0, 0, 0, 0, 0, 1,…
## $ sqft__887_1100                              <dbl> 0, 0, 0, 1, 0, 0, 0, 1, 0,…
## $ sqft__1100_1500                             <dbl> 0, 0, 1, 0, 1, 1, 0, 0, 0,…
## $ sqft__1500_Inf                              <dbl> 1, 0, 0, 0, 0, 0, 1, 0, 0,…

# Step 2: Correlate
data_corr_tbl <- data_binarized_tbl %>%
    correlate(price__8.07868822922987_Inf)

data_corr_tbl

## # A tibble: 85 × 3
##    feature bin                               correlation
##    <fct>   <chr>                                   <dbl>
##  1 price   8.07868822922987_Inf                    1    
##  2 city    san_francisco                           0.389
##  3 county  san_francisco                           0.389
##  4 price   -Inf_7.52294091807237                  -0.342
##  5 price   7.80384330353877_8.07868822922987      -0.330
##  6 price   7.52294091807237_7.80384330353877      -0.328
##  7 sqft    1500_Inf                                0.324
##  8 beds    -Inf_2                                 -0.254
##  9 beds    3_Inf                                   0.241
## 10 sqft    -Inf_887                               -0.240
## # ℹ 75 more rows

# Step 3: Plot
data_corr_tbl %>%
    plot_correlation_funnel()

## Warning: ggrepel: 69 unlabeled data points (too many overlaps). Consider
## increasing max.overlaps

Build Models

Split data

#data <- data %>% filter(beds > 0)
 data <- data %>% sample_n(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 [67/8]> Fold01
##  2 <split [67/8]> Fold02
##  3 <split [67/8]> Fold03
##  4 <split [67/8]> Fold04
##  5 <split [67/8]> Fold05
##  6 <split [68/7]> Fold06
##  7 <split [68/7]> Fold07
##  8 <split [68/7]> Fold08
##  9 <split [68/7]> Fold09
## 10 <split [68/7]> Fold10

library(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(6804)
## xgboost_tune <-
##   tune_grid(xgboost_workflow, resamples = stop("add your rsample object"), grid = stop("add number of candidate points"))

# Specify Recipe
# Use either step_tf or step_tfidf
xgboost_recipe <- 
  recipe(formula = price ~ ., data = data_train) %>%
    recipes::update_role(post_id, new_role = "id variables") %>%
    step_tokenize(title) %>%
    step_tokenfilter(title, max_tokens = 100) %>%
    step_tfidf(title) %>%
    step_other(nhood, city) %>%
    step_dummy(nhood, city, county, one_hot = TRUE) %>%
    step_log(sqft, beds, baths)

xgboost_recipe %>% prep() %>% bake(new_data = NULL) %>% glimpse()

## Rows: 75
## Columns: 120
## $ post_id                  <fct> 6359740472, 4943342358, pre2013_54450, 505474…
## $ beds                     <dbl> 1.0986123, 0.6931472, 0.6931472, 0.0000000, 0…
## $ baths                    <dbl> 0.6931472, 0.6931472, 0.0000000, 0.0000000, 0…
## $ sqft                     <dbl> 7.215240, 7.167038, 6.684612, 6.476972, 7.003…
## $ price                    <dbl> 8.571303, 8.476371, 7.600902, 7.673223, 8.496…
## $ tfidf_title_1            <dbl> 0.00000000, 0.12330527, 0.00000000, 0.6165263…
## $ tfidf_title_1.5          <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tfidf_title_1100ft       <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tfidf_title_1150ft       <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_1200         <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tfidf_title_1200ft       <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tfidf_title_1ba          <dbl> 0.0000000, 0.0000000, 0.3718128, 0.0000000, 0…
## $ tfidf_title_1br          <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_1st          <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_2            <dbl> 0.00000000, 0.25055259, 0.17896614, 0.0000000…
## $ tfidf_title_2.5          <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_2.5ba        <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tfidf_title_22           <dbl> 0.0000000, 0.2433772, 0.0000000, 0.0000000, 0…
## $ tfidf_title_2ba          <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_2bath        <dbl> 0.3960841, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_2bedroom     <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_2br          <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_3            <dbl> 0.0000000, 0.1194506, 0.0000000, 0.0000000, 0…
## $ tfidf_title_30p          <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_3ba          <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_3bd          <dbl> 0.5215226, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_3bed         <dbl> 0.000000, 0.000000, 0.000000, 0.000000, 0.000…
## $ tfidf_title_3br          <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_4            <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_4750         <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_4bd          <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_4br          <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_5            <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tfidf_title_6            <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_8            <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_900ft        <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tfidf_title_a            <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_amazing      <dbl> 0.0000000, 0.2433772, 0.0000000, 0.0000000, 0…
## $ tfidf_title_amp          <dbl> 0.5215226, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_and          <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_apartment    <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_apt          <dbl> 0.5215226, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_area         <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_at           <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_available    <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_ba           <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_balcony      <dbl> 0.0000000, 0.2433772, 0.0000000, 0.0000000, 0…
## $ tfidf_title_bath         <dbl> 0.00000000, 0.08047286, 0.00000000, 0.2011821…
## $ tfidf_title_bathroom     <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tfidf_title_baths        <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tfidf_title_bd           <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_beautiful    <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_bed          <dbl> 0.0000000, 0.1233053, 0.0000000, 0.0000000, 0…
## $ tfidf_title_bedroom      <dbl> 0.0000000, 0.0000000, 0.0000000, 0.3301669, 0…
## $ tfidf_title_bonus        <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tfidf_title_br           <dbl> 0.0000000, 0.0000000, 0.3515442, 0.0000000, 0…
## $ tfidf_title_by           <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tfidf_title_close        <dbl> 0.4654424, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_condo        <dbl> 0.0000000, 0.2172064, 0.0000000, 0.0000000, 0…
## $ tfidf_title_cupertino    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tfidf_title_downtown     <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_duplex       <dbl> 0.0000000, 0.0000000, 0.4654424, 0.0000000, 0…
## $ tfidf_title_floor        <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_for          <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_garage       <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tfidf_title_great        <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_hide         <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_home         <dbl> 0.3515442, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_house        <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_huge         <dbl> 0.0000000, 0.2172064, 0.0000000, 0.0000000, 0…
## $ tfidf_title_in           <dbl> 0.00000000, 0.00000000, 0.21195265, 0.0000000…
## $ tfidf_title_large        <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tfidf_title_laundry      <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_location     <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tfidf_title_new          <dbl> 0.0000000, 0.0000000, 0.0000000, 0.5430161, 0…
## $ tfidf_title_nice         <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_of           <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_off          <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_on           <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_one          <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tfidf_title_parking      <dbl> 0.0000000, 0.1988769, 0.0000000, 0.0000000, 0…
## $ tfidf_title_posting      <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_private      <dbl> 0.0000000, 0.0000000, 0.4654424, 0.0000000, 0…
## $ tfidf_title_remodeled    <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_rent         <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_restore      <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_san          <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_schools      <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_single       <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_spacious     <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_street       <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_the          <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tfidf_title_this         <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_to           <dbl> 0.3718128, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_townhome     <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tfidf_title_townhouse    <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tfidf_title_two          <dbl> 0.000000, 0.000000, 0.000000, 0.000000, 0.000…
## $ tfidf_title_unit         <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_view         <dbl> 0.0000000, 0.0000000, 0.4654424, 0.0000000, 0…
## $ tfidf_title_views        <dbl> 0.0000000, 0.1735126, 0.0000000, 0.0000000, 0…
## $ tfidf_title_w            <dbl> 0.0000000, 0.1640539, 0.0000000, 0.0000000, 0…
## $ tfidf_title_with         <dbl> 0.0000000, 0.0000000, 0.0000000, 0.3566777, 0…
## $ tfidf_title_이           <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ tfidf_title_地图         <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ tfidf_title_月           <dbl> 0.0000000, 0.0000000, 0.0000000, 0.0000000, 0…
## $ nhood_san.jose.central   <dbl> 0, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, …
## $ nhood_SOMA...south.beach <dbl> 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ nhood_other              <dbl> 1, 1, 1, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, …
## $ city_san.francisco       <dbl> 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, …
## $ city_san.jose            <dbl> 0, 0, 0, 1, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, …
## $ city_other               <dbl> 1, 0, 1, 0, 0, 0, 1, 1, 1, 0, 1, 1, 0, 0, 1, …
## $ county_alameda           <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, …
## $ county_contra.costa      <dbl> 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, …
## $ county_marin             <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, …
## $ county_san.francisco     <dbl> 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, …
## $ county_san.mateo         <dbl> 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ county_santa.clara       <dbl> 0, 0, 1, 1, 0, 1, 1, 0, 0, 1, 1, 0, 0, 0, 0, …
## $ county_santa.cruz        <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ county_solano            <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …
## $ county_sonoma            <dbl> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …

xgboost_spec <- 
    boost_tree(trees = tune(), min_n = tune(), tree_depth = 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(37369)
doParallel::registerDoParallel()
xgboost_tune <-
  tune_grid(xgboost_workflow, 
            resamples = data_cv, 
            grid = 5)

Evaluate Models

tune::show_best(xgboost_tune, metric = "rmse")

## # A tibble: 5 × 10
##   trees min_n tree_depth learn_rate .metric .estimator  mean     n std_err
##   <int> <int>      <int>      <dbl> <chr>   <chr>      <dbl> <int>   <dbl>
## 1  1147    24          5    0.0481  rmse    standard   0.387    10  0.0373
## 2   691    11          8    0.00530 rmse    standard   0.401    10  0.0377
## 3  1681    37         11    0.108   rmse    standard   0.436    10  0.0375
## 4   235    28         12    0.0114  rmse    standard   0.645    10  0.0452
## 5  1384     7          1    0.00131 rmse    standard   1.29     10  0.0577
## # ℹ 1 more variable: .config <chr>

# Update the model by selecting the best hyperparameters
xgboost_fw <- tune::finalize_workflow(xgboost_workflow,
                        tune::select_best(xgboost_tune, metric = "rmse"))

# Fit the model on the entire traning 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       0.512 Preprocessor1_Model1
## 2 rsq     standard       0.223 Preprocessor1_Model1

tune::collect_predictions(data_fit) %>%
    ggplot(aes(price, .pred)) +
    geom_point(alpha = 0.3, fill = "midnightblue") +
    geom_abline(lty = 2, color = "gray50") +
    coord_fixed()

Code along 1: SF Rental Market

Julius Mondschein

2024-09-11

Import Data

Explore Data

Build Models

Evaluate Models