TakeHomeEx2

Author

Joshua TING

Published

April 28, 2024

Modified

May 29, 2024

The Private Real Estate Market: DataVis Make Over Edition

[Source: The Interlace] (https://www.dezeen.com/2014/10/07/ole-scheeren-the-interlace-important-prototype-housing-waf-2014/)

Introduction

1.1 Setting The Scene

80% of Singapore residents reside in public housing while 20% resides in private housing. In the private housing market, land ownership is divided into two parts: land and strata. For landed property, the plot of land belongs to the owner while in strata property, the plot of land is jointly owned by the legal owners in the same development (The Business Times, 2022). A detached house, commonly known as bungalow, semi-detached house and terrace house may possess either be landed or strata titled.

1.2. Objective

In this exercise, we aim to apply different data visualisation design practices and principles in improving on the Take-Home Exercise 1’s output of a fellow coursemate

Preparation

In this section, we will be preparing our R environment and data set.

2.1 Packages

The following functions will be loaded using ‘pacman:p_load()’ in R Packages to facilitate the data preparation and analysis process.

Display Code 
pacman::p_load(tidyverse,dplyr, ggthemes,colorspace,ggiraph,
               plotly,patchwork,lubridate, 
               ggrepel,ggdist)

2.2 Data

Display Code 
q5data <- read_csv("data/ResidentialTransaction20240414220633.csv")

A total of 4,902 rows and 21 columns. If we want to look at the attributes of the data, we can use glimpse() function as seen in next section 2.3.

2.3 Glimpse

Display Code 
glimpse(q5data)
Rows: 4,902
Columns: 21
$ `Project Name`                <chr> "THE LANDMARK", "POLLEN COLLECTION", "SK…
$ `Transacted Price ($)`        <dbl> 2726888, 3850000, 2346000, 2190000, 1954…
$ `Area (SQFT)`                 <dbl> 1076.40, 1808.35, 1087.16, 807.30, 796.5…
$ `Unit Price ($ PSF)`          <dbl> 2533, 2129, 2158, 2713, 2453, 2577, 838,…
$ `Sale Date`                   <chr> "01 Jan 2024", "01 Jan 2024", "01 Jan 20…
$ Address                       <chr> "173 CHIN SWEE ROAD #22-11", "34 POLLEN …
$ `Type of Sale`                <chr> "New Sale", "New Sale", "New Sale", "New…
$ `Type of Area`                <chr> "Strata", "Land", "Strata", "Strata", "S…
$ `Area (SQM)`                  <dbl> 100.0, 168.0, 101.0, 75.0, 74.0, 123.0, …
$ `Unit Price ($ PSM)`          <dbl> 27269, 22917, 23228, 29200, 26405, 27741…
$ `Nett Price($)`               <chr> "-", "-", "-", "-", "-", "-", "-", "-", …
$ `Property Type`               <chr> "Condominium", "Terrace House", "Apartme…
$ `Number of Units`             <dbl> 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1…
$ Tenure                        <chr> "99 yrs from 28/08/2020", "99 yrs from 0…
$ `Completion Date`             <chr> "Uncompleted", "Uncompleted", "Uncomplet…
$ `Purchaser Address Indicator` <chr> "Private", "N.A", "HDB", "N.A", "Private…
$ `Postal Code`                 <chr> "169878", "807233", "469657", "118992", …
$ `Postal District`             <chr> "03", "28", "16", "05", "21", "21", "28"…
$ `Postal Sector`               <chr> "16", "80", "46", "11", "59", "58", "79"…
$ `Planning Region`             <chr> "Central Region", "North East Region", "…
$ `Planning Area`               <chr> "Outram", "Serangoon", "Bedok", "Queenst…

As evidenced, glimpse() function indicates the variable, the type of variable (chr, dpl, num) and the subvariables within the variable.

2.4 Converting dbl (double precision) to num (number)

As there is no decimal points in this data set, the variable “Transacted Price ($)” will be changed to number instead of double precision.

Display Code 
# Convert "Transacted Price ($)" to numeric
q5data$`Transacted Price ($)` <- as.numeric(gsub("[^0-9.]", "", q5data$`Transacted Price ($)`))

2.5 Calculating mean of each property type

The average of each property type such as Condominium, Terrace House will be calculate in facilitating the data makeover process in the later sections.

Tip

The More You Know

The %>% operator, pronounced as “then”, is part of the magrittr package in R. It’s used for piping, which allows you to perform a sequence of operations on data without nested function calls, making your code more readable and concise.

Display Code 
mean_prices <- q5data %>%
 group_by(`Property Type`, `Type of Area`) %>%
 summarise(mean_price = mean(`Transacted Price ($)`), .groups = "drop")

Critic

The critic will be based Figure 2 and the article “Data Visualization: Clarity or Aesthetics” acts as a scaffold of assessment. The diagram of the coordinate system below will served as an overall assessment of the visualisation.

Figure 1: Data Visualization: Clarity vs. Aesthetics

Figure 2: The Product

3.1. Overall

The design is elegant; is not bombarded with mouthful of information. It allows readers to examine the median transacted prices of each propery type in the type of area at a glance. The background colour is easy on the user’s eyes; not too bright nor contrasting. Good consideration.

3.2. Clarity

In the context of Jones (2012), the author defined clarity as “how quickly and effectively it imparts to the audience an accurate understanding of some fundamental truth about the real world”. The two metrics that will be used in this section as per follows:

1. At one glance, does it tell a story?

2. Upon examination, Is it effective in telling me a story within the visulisation?

Comments:

  1. The dots that represents the median price of each property type is rather small. Audience may not effectively picked up the median price of the size of the dot

    Solution: To increase dotsize

  2. The lack of the actual median price being displayed beside the dot does not tell the audience

    Solution: To indicate the average price beside the dot

  3. The use of median prices might be one of the market’s indicator for the transactions prices. However, average transacted prices is also an important market indicator despite its sensitivity to outlier transactions. Stakeholders such as potential buyers do want to know the average transacted price too as it may indicate they potentially need to purchase the property.

    Solution: Box & Whisker plot will be introduced instead as the audience is able to clearly visualise the distribution of the transacted price. The box & whisker plot will reflect the following: 1. Lower & Upper Quantile 2. Minimum & Maximum Value 3. Outliers 4. Average price 5. Median price

3.3. Aesthetic

Jones (2012) dictates that aesthetics should only be discussed once clarity has been achieved. In the previous section, shortfalls has been discussed alongside proposed solutions. Therefore, upon achieving clarity, aesthetics can be explored.

  1. The inverted words of the property type makes it difficult for the audience to comprehend each column. Solution:

  2. Lack of title in this visualisation - Audience might not know which quarter or year this data was obtained from

Product Realisation

4.1 Ideation

A rough sketch of the box and whiskers plot has been made in facilitating the ideation process.

4.2 Blackboard

Attempt 1 - Clarity:

The data was populated into a box and whiskers plot and the below plot was generated. Noticed that the plots appear compressed especially in Strata due to the presence of outlier transaction prices seen in the Detached House. Thus, in this instance, it is practical to remove these 2 outliers transactions.

Display Code 
ggplot(q5data, aes(x = `Transacted Price ($)`, y = `Property Type`)) +
  geom_boxplot() +
  scale_x_continuous(limits = c(500000, 40500000), breaks = seq(500000, 40500000, 5000000), labels = scales::dollar_format(prefix = "$")) +
  facet_wrap(~`Type of Area`, ncol = 1, labeller = labeller(`Type of Area` = c(Strata = "Strata"))) +
  labs(x = "Transacted Price ($)", y = NULL) +
  theme_minimal(base_size = 8) +
  theme(axis.text = element_text(size = 8),
        strip.text = element_text(size = 10),
        legend.position = "none")

Attempt 2 - Clarity:

By limiting the x axis to $20,500,000, the two outlier transactions were removed. At a first glance, it is clear that there are no transactions of Executive Condomimium and Condominium within the Strata type. Additionally, there is only one transaction for ‘Apartment’ in Land and ‘Detached House’ in Strta.

Display Code 
ggplot(q5data, aes(x = `Transacted Price ($)`, y = `Property Type`)) +
  geom_boxplot() +
  scale_x_continuous(limits = c(500000, 20500000), breaks = seq(500000, 20500000, 5000000), labels = scales::dollar_format(prefix = "$")) +
  facet_wrap(~`Type of Area`, ncol = 1, labeller = labeller(`Type of Area` = c(Strata = "Strata"))) +
  labs(x = "Transacted Price ($)", y = NULL) +
  theme_minimal(base_size = 8) +
  theme(axis.text = element_text(size = 8),
        strip.text = element_text(size = 10),
        legend.position = "none")

Attempt 3 - Clarity:

An attempt has been made to allow the box and whiskers plot to be posiitoned vertically instead of horizontally as it may faciliate in assessing the trasanced price from lowest to the highest. However, the plot appeared compressed and might be secondary to orientation of the diagram. Hence the decision was made to retain the original plot as seen in Attempt 2.

Display Code 
ggplot(q5data, aes(x = `Property Type`, y = `Transacted Price ($)`)) +
  geom_boxplot() +
  scale_y_continuous(limits = c(500000, 20500000), breaks = seq(500000, 20500000, 5000000), labels = scales::dollar_format(prefix = "$")) +
  facet_wrap(~`Type of Area`, ncol = 1, labeller = labeller(`Type of Area` = c(Strata = "Strata"))) +
  labs(x = NULL, y = "Transacted Price ($)") +
  theme_minimal(base_size = 8) +
  theme(axis.text = element_text(size = 8),
        strip.text = element_text(size = 10),
        legend.position = "none")

Attempt 4 - Clarity:

In ensuring the clarity, the average transacted prices for each property type has been added alongside the actual average transacted figure and in facilitating the flow of vision, the color “blue” has been used.

Display Code 
ggplot(q5data, aes(x = `Transacted Price ($)`, y = `Property Type`)) +
  geom_boxplot() +
  stat_summary(fun.y = mean, geom = "point", shape = 18, size = 2, color = "blue") +
  geom_text(data = mean_prices, aes(label = scales::dollar_format()(mean_price), y = `Property Type`, x = mean_price), vjust = 2, hjust= -0.5, size = 2.5, color="blue") +
  scale_x_continuous(limits = c(500000, 20500000), breaks = seq(500000, 20500000, 5000000), labels = scales::dollar_format(prefix = "$")) +
  facet_wrap(~`Type of Area`, ncol = 1, labeller = labeller(`Type of Area` = c(Strata = "Strata"))) +
  labs(x = "Transacted Price ($)", y = NULL) +
  theme_minimal(base_size = 8) +
  theme(axis.text = element_text(size = 8),
        strip.text = element_text(size = 10),
        legend.position = "none")

Attempt 5 - Aesthetics:

Each individual property type has been assigned a colour to faciliate viewing. The colour is consistent in both Land and Strata. For example: The use of red colour is standardised for Terrace House in Land and Strata type. However, visualisation as a whole, appears bright which not not aid viewing. Hence the last attempt will change the background colour.

Display Code 
#Vector of colors for each Property Type
property_type_colors <- c("Terrace House" = "red", "Semi-Detached House" = "blue", 
                           "Detached House" = "green", "Condominium" = "purple", 
                           "Executive Condominium" = "orange")

#Boxplot
ggplot(q5data, aes(x = `Transacted Price ($)`, y = `Property Type`, color = `Property Type`)) +
 geom_boxplot() +
 stat_summary(fun.y = mean, geom = "point", shape = 18, size = 2) +
 geom_text(data = mean_prices, aes(label = scales::dollar_format()(mean_price), y = `Property Type`, x = mean_price), vjust = 2, hjust = -0.5, size = 2.5) +
 scale_color_manual(values = property_type_colors) +
 scale_x_continuous(limits = c(500000, 20500000), breaks = seq(500000, 20500000, 5000000), labels = scales::dollar_format(prefix = "$")) +
 facet_wrap(~`Type of Area`, ncol = 1, labeller = labeller(`Type of Area` = c(Strata = "Strata"))) +
 labs(x = "Transacted Price ($)", y = NULL) +
 theme_minimal(base_size = 8) +
 theme(axis.text = element_text(size = 8),
       strip.text = element_text(size = 10),
       legend.position = "none")

Final Product

5.1 Before

5.2 Makeover

Display Code 
#Vector of colors for each Property Type
property_type_colors <- c("Terrace House" = "red", "Semi-Detached House" = "blue", 
                           "Detached House" = "darkgreen", "Condominium" = "purple", 
                           "Executive Condominium" = "orange")

#Boxplot
ggplot(q5data, aes(x = `Transacted Price ($)`, y = `Property Type`, color = `Property Type`)) +
 geom_boxplot() +
 stat_summary(fun.y = mean, geom = "point", shape = 18, size = 2) +
 geom_text(data = mean_prices, aes(label = scales::dollar_format()(mean_price), y = `Property Type`, x = mean_price), vjust = 3, size = 2.2, fontface = "bold") +
 scale_color_manual(values = property_type_colors) +
 scale_x_continuous(limits = c(500000, 20500000), breaks = seq(500000, 20500000, 5000000), labels = scales::dollar_format(prefix = "$")) +
 facet_wrap(~`Type of Area`, ncol = 1, labeller = labeller(`Type of Area` = c(Strata = "Strata"))) +
 labs(x = "Transacted Price ($)", y = NULL, title = "Transacted Prices of Private Residential Market in Q12024") +
 theme_minimal(base_size = 8) +
 theme(axis.text = element_text(size = 8),
       strip.text = element_text(size = 10),
       legend.position = "none",
       plot.background = element_rect(fill = "lightblue"))

Conclusion

Overall, the makeover enhanced the clarity and aesthetics of the former visulisation. It allows the audience to visualise almost the full spectrum of the data alongside its statistics.

6.1. Remarks

a\. The visualisation might look cluttered as two diagrams of box and whiskers plot were chunked into one. However, the comparison was necessary in visualing the transacted prices using the same x-axis.

b\. The numbers of the average transacted price labelled within each box and whisper plot were small and some of the numbers were cut out such as Apartment in both Land and Strata.

c\. Two outliers were eradicated as the box and whiskers plot appeared very narrowed. Thus the full spectrum of the outliers was not shown.

Post Submission Add-On

Coming Soon.

References

Jones, B. (2012). Data Visualization: Clarity or Aesthetics. Retrieved from https://dataremixed.com/2012/05/data-visualization-clarity-or-aesthetics/

The Business Times. (2022). Landed home prices set to stay firm, if not trend upwards. Retrieved from https://www.businesstimes.com.sg/property/landed-home-prices-set-stay-firm-if-not-trend-upwards