Difference between revisions of "ISSS608 2016-17 T1 Assign1 XU Qiuhui"
Line 6: | Line 6: | ||
=Data= | =Data= | ||
== Data Acquisition == | == Data Acquisition == | ||
− | |||
== Dataset Variables == | == Dataset Variables == |
Revision as of 11:34, 29 August 2016
Contents
Abstract
<TBD>
Problem and Motivation
Data
Data Acquisition
Dataset Variables
In the data.gov.sg dataset, a quick review of the variables reveals no missing data.
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 and Rationale |
---|---|
Month | A new variable "Date" was created to parse the YYYY-MM format of the Month variable. The formula DATEPARSE('yyyy-MM',[Month]) was used in the calculated field. |
Town | The variable is set to a State/Province geographical role for purpose of applying it to filled maps in the analysis. |
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.
Data Analysis
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>
Most popular: 4 Room, 3 Room.
Visualisation rationale: <TBD>
Lower floors are sold more. House owners prefer to keep higher floors?
Visualisation rationale: <TBD>
Distribution of the resale public housing prices in 2015
No surprises here. Median prices are higher in mature estates and areas near the city centre such as Central area, Queenstown.
Visualisation rationale: Median avoids outliers. <TBD>
Since primary flats sold are 4 Room, let's focus on 4 Room. Focus on families on a budget.
Visualisation rationale: <TBD>
No surprises on the estates.
Visualisation rationale: <TBD>
Comparison of patterns of the first-half of 2016 with the patterns of 2015
Visualisation rationale: <TBD>
Visualisation rationale: <TBD>
Visualisation rationale: <TBD>
Infographics
Tools Utilised
- SAS JMP 12 – for initial data exploration and analysis
- Tableau 10.0 – for charting
- Microsoft Powerpoint – for Infographics layout
Results