ISSS608 2016-17 T1 Assign1 Meenakshi
Abstract
Public housing in Singapore is governed by the Housing and Development board. HDB flats are home to over 80% Singaporeans and hence it is important for all the stake holders to understand the supply and price trends for public housing. This will greatly influence the planning in future.
Problem and motivation
The study aims at -
- Understanding the share of supply of resale public housing in 2015.
- Studying the distribution of resale prices with respect to important dimensions such as town and flat type.
- Compare the patterns for 2015 and first half of 2016.
- Note key observations and inferences through visual analytics. The goal is to note a minimum of 3 observations in each case.
Tools Used
Tableau version 10.0
Approaches
- Data collection and cleaning:
The housing data for 2015 and 2016 is available at https://data.gov.sg .The data is clean and ready for analysis. It was necessary to format few fields and data types to be ready to use with the tool.
- Understanding the data:
The various dimensions used in the data set needs to be recorded. This can be easily achieved with appropriate visualization tools. Then we make a note on the different parameters we wish to study and the suitable graphical representation for each case. For instance, Histograms and Box plots could be used to study the Distribution of continuous data types such as prices.
- Analysis:
The data set is imported to the tool and Visual infographics designed for each case. With the graphs and various statistical parameters we note our key observations and compare patterns.
Results
The results of the study are shown in the infographics below. The observations will enable stake holders to see the past and current trends for housing supply and prices. Business decisions can be influenced positively by studying the demand and supply.
Public Housing supply and resale price distribution for 2015/2016
The results are summarized visually in the below infographics
The graph below shows the number of resale applications registered in 2015. We can see that in Q2 there is a surge in supply for commonly transacted flat types.
The number of units transacted can be compared for the first half of 2015 and 2016. The stacked bars show the shares by room type for Q1 and Q2. The common trend is that 4 room, 3 room and 5 room units contribute mostly to the totals. The trend seems to be repeating in both the years.
Median Resale Price, Count of transactions by Town
The Histograms show a distribution of number of resale transactions for both 2015 and 2016 town wise. However for 2016 only Q1Q2 data is included. Clearly the distribution seen in 2015 is repeating in 2016.
On similar lines, the price distribution can be compared. The graphs clearly show the trends repeating. We can mark down the prime locations that are highly priced compared to few others.
Detailed resale price Distribution by Town
The detailed distribution of prices by town can be analysed with the Box and Whisker's plot. We see that those locations where fewer units are being transacted, rather contribute more to the %total of prices transacted. For instance Bukit Timah and Central area are on top in the graph. Also the price range is highly varying in these locations.
Where as in few locations such as Yishun, Pasir Ris the units are more closely priced to median price.
One key point to note is that most high priced units, that is the outliers are the Executive type flats. Hence looking at the median prices alone might mislead to conclude that 4 room, 5 room unit resale prices are also very high.
Major contributors to Percent total of resale price
The Pareto chart below shows that 80% of the total transacted resale price is contributed by the sales of 4 Room, 5 Room, and 3 Room type flats.
This can be correlated to the fact that the supply of 4 Room and 3 Room units are higher compared to others.
However a pareto distribution of %total of prices by town type can be added to understand the distribution. Alternately, a distribution of Flat Types by Town, helps understand which towns are contributing most to the total of resale prices transacted.
Two pareto charts each for 2015 and 2016 are added and we see the trends repeating.
