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

Import Data

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

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

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 pos-skewed distributions
  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 avg rent per word
  group_by(word) %>%
  summarise(price = mean(price),
            n = n()) %>%
  ungroup() %>%
    
# Filter for words appearing > 10 times and remove digits
  filter(n > 10, !str_detect(word, "\\d")) %>%
  slice_max(order_by = price, n = 20) %>%
  
  ggplot(aes(price, fct_reorder(word, price))) +
  geom_point() +
  labs(y = "words in title")

EDA shortcut

# 1. Binarize the data (creates the input)
data_binarized_tbl <- data %>%
    select(-post_id, -title) %>%
    binarize()

# 2. Run Correlation (Fix: Use 'data_binarized_tbl' as the input)
data_corr_tbl <- data_binarized_tbl %>%
    correlate(target = price__8.07868822922987_Inf)

# 3. Plot the result
data_corr_tbl %>%
    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 per session.
## 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 per session.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.
## Warning: ggrepel: 69 unlabeled data points (too many overlaps). Consider
## increasing max.overlaps

Preprocess Data

Build Data

Evaluate Data

Make Predictions