ISSS608 2016-17 T1 Assign1 CHIA Yong Jian

From Visual Analytics and Applications
Revision as of 06:58, 29 August 2016 by Yongjian.2015 (talk | contribs)
Jump to navigation Jump to search

Abstract

<TBD>

Problem and Motivation

The public housing resale market is a closely watched market. A quick search on news revealed that the public housing resale market movement is watched closely almost on a monthly basis, especially for indicators hinting the stability or trend of the prices. These are crucial information not just for real estate companies, but also buyers and sellers of resale flats. This analysis is focused on providing relevant information that can aid buyers and sellers of resale flats to make informed decisions when deciding to buy or sell a flat.

  • For buyers, they include newly-wed couples that may urgently need a house for their family needs. Knowing the prices of resale flats by areas can allow them to decide on a suitable area for house-hunting depending on their budget and preferences for amenities in the area
  • For sellers, they are existing house owners that might be watching the market to understand the general market trends to decide on a good price to sell their houses.

We assume that the buyers are sellers are the general population where they might be untrained in statistics. Hence, charts would be generally designed from an approach that are visually attractive without unnecessary fluff, and requires little time for them to understand without requiring statistical knowledge.

Data

Data Acquisition

From Data.gov.sg

The dataset used to answer the above questions can be found from Data.Gov.Sg: Resale Flat Prices . There are two datasets available on the webpage. We will be using the resale flat prices data that includes data for the year 2015 to first half 2016, to answer the questions required.

From HDB Website

Supplementary data from HDB website is used to understand better the resale flat supply by the 23 towns and 3 estates planned by HDB, by normalizing the resale flats sold by the total flat supply in the area. The data is retrieved from HDB Towns: Your Home .

Dataset Variables

In the data.gov.sg dataset, a quick review of the variables reveals no missing data.

Missing Data Checks in SAS JMP


The dataset includes the following variables:

Variable Name Description Example Levels/Values
Month The 4-digit year and 2-digit month of the transaction From “2012-03” to “2016-06”
Town Includes the 23 towns and 3 estates under HDB planning ANG MO KIO
BEDOK
BISHAN
Flat Type The flat type sold. More information can be gotten from HDB: Types of Flats 1 ROOM
5 ROOM
EXECUTIVE
Block The block number of where the flat was sold. 99C
99B
977
Street Name The street name in the town or estate where the flat was sold. ZION RD
YUNG SHENG RD
WHAMPOA WEST
Storey Range The storey of the flat sold. Ranges from first floor to highest fifty-first floor. Floors are binned to groups of 3 to 5, and may include overlaps 01 TO 03
25 TO 27
43 TO 45
Floor Area Sqm The floor area of the flat sold. Ranges from 31 to 280 square metres. 45
74
92
Flat Model The model type of the flat sold. There are 19 types. Details of the different flat models can be view on a website providing comprehensive information: HDB History and Floor Plan Evolution Adjoined Flat
Model A
New Generation
Lease Commencement The starting year of the flat lease, typically a 99-year period 1966
2004
2013
Resale Price The transacted price of the flat sold. $195,000
$515,000

$1,088,888

Data Modelling

The following are changes made before charting was done on Tableau: <TBD>

Variable Changes Made Rationale
Example Example Example
Example Example Example

Approaches

General Colour Scheme for Charts

A consistent colour scheme was used in the charts to retain a sense of familiarity with the readers and reduce confusion. Several colour schemes was experimented for use across the charts. The final colour scheme selected deliberately used contrasting colours. Depending on the chart used, either 3 steps or 5 steps colours was used to allow readers to easily understand the concept of "high, medium, low" without further explanation.
Selected Colour Scheme

Data Analysis

Share of the resale public housing supply in 2015

Number of Resale Transactions by town

An overlay of the number of resale transactions by each HDB town or estate on the Singapore Map reveals that towns with the most resale flat transactions are from Jurong West, Tampines, Bedok, Sengkang and Woodlands. It can be observed that most of these areas are well built-up, and includes amenities such as shopping centres.

Visualisation rationale: <TBD>


Tools Utilised

  1. SAS JMP 12 – for initial data exploration and analysis
  2. Tableau 10.0 – for charting
  3. Microsoft Powerpoint – for Infographics layout




Results