I’ve created this fusion table to highlight the population density in Ireland. This involved merging 2 tables together on Google Fusion. I downloaded the kml file for Ireland’s geographic locations and then merged it with the population of Ireland KML table. I then had to create the population bands which highlighted the higher and lower population densities in difference colours. When choosing the population density bands, I first checked what the populations of each county were, so that I could create population density bands that would reflect the information in a meaningful way.
I noticed that less than 100,000 was common to a few counties, so I picked this as a starting population band. Then I noticed that there were quiet a few counties in the 100,000 – 200,000 range, so I then picked this as another population range. After that I decided that it was important to highlight the 200,000 – 500,000 range as there were 2 counties that fell into this range. After this, I felt that it was important to highlight the 500,000 – 750,000 range which represents Cork, which I though was appropriate as it is the 2nd biggest city in Ireland, and then of course I represented Dublin in blue on it’s own as there are no other counties in Ireland with as big a population as Dublin.
This information could be used to highlight the need for roads, the need for development in these areas?
One thing I think this might highlight is that the coastal locations seem to be quiet popular for settlement and development and the midlands were not as popular for settlement. This could be for multiple reasons such as initially people would have settled near shipping ports so that they could get jobs involved with shipping and there would have been more of a social scene near shipping ports and this trend continued. My own opinion would be that there is a premium in house prices for living by the sea because people like to be near the sea for scenery, walks and entertainment.
Another interesting point why people do not settle as much in the midlands would be that a lot of the land would be used for farming and maybe not so good for building houses on. Some people prefer the quiet and cheaper life in the midlands, but it seems that more people are attracted to the busy social scene of big cities such as Dublin.
Going forward, as Dublin house prices are set to soar again, it would be interesting to monitor how the population of counties outside Dubin react to the house price increases?. Will the increase in house prices cause a population surge in other counties?
An interesting study would be to get further population settlement locations to highlight the most attractive locations for historic people to settle….such as near the coast for fishing, away from flood plains, good farming land nearby, near shipping ports.
Here is the link to look at my fusion table in more detail and you could amend the structure if you wanted to amend the focus of the table
R Code School
I found R to be a good introduction into programming and how the programming language works. We started off with setting variables and then using these variables in calculations. I could quickly see that setting a variable as a text name and then using that variable name in later functions could be really useful when working with big data. We then looked at the R HELP function which was easy to use and could be called upon at anytime during programming to assist with the setup of a functions.
We then moved on to look at listing files from a particular directory. This was really interesting for me as I could see how a large number of files in the same folder location, could easily be listed and pulled together to form huge data sets. I then realised how powerful R could be as opposed to a programme like Excel where this functionality would not be available and a very manual process would be needed to gather data from multiple files.
We then moved on to setting up vectors and then plotting them on a graph. It was very simple to set the vectors up and then plot them and I could see how this could be very useful when working with data. It helps you to look at the data and determine how you would like to use it in a logical way and then plot it on a simple graph for analytics with the choice of allowing for values that are not valid.
We then moved on to setting up matrices with multiple rows /columns and then assigning values to each row / column by setting values. It quickly became very clear that you could easily setup a matrix with a high number of rows and columns and this information could quickly and easily be recalled from the matrix by using the correct formula. This huge dataset could then be easily plotted on a graph using the R graph functions such as the image(volcano) function which creates a heat map of the matrix.
Next up was using R to calculate statistical mean, mode and standard deviation for datasets. At this stage, it is starting to become very clear how logical R is and it’s ease of use to calculate complicated formulas and then plot them on to graphs. As soon as you get familiar with the process of inputting your information, then it becomes very simple to work with this information to analyse and plot graphs. I will definitely keep this application in mind for future projects that I may have and I would be very interested to take a look at some of the more advanced functions available to R users.
Big Data and it’s use in financial trading…
Big Data is the term given to manage large datasets that consist of unstructured, semi-structured and structured data from many sources that cannot be interpreted by traditional database methods or tools. Big Data was first broken down by Gartner into the 3 V’s which consist of Volume, Velocity and Variety. These 3 V’s are the main problems faced when trying to trade the financial markets, so I decided to see how Big Data interacted / was applied to Financial Trading, because the volume of transactions that move the market prices is very high, the speed at which prices change is so fast and the relevant information and it’s sources are so vast, so it would make sense that Big Data should be applied to Financial Trading to reduce risk and hopefully have a profitable outcome.
As I investigated, it seems that Big Data predictive analytics itself is evolving within Financial Trading – from the unethical high frequency algorithmic trading over the past few years which is based on the last 5 minutes of market prices/ trading. This has in recent times caused market crashes /market gaps in the past, but it seems that Big Data has helped to move away from this kind of Financial Trading, on to a more long term algorithmic trading which is based on Analysing the long term structure / trends in the market, finding predictable patterns, and creating technology derived forecasts.
By analysing the outcomes of a multitude of trades, systematic trend patterns often emerge, which helps to better understand the past market trends, aswell as predict realistic market forecasts for the future. There are three different patterns the market can alternate between – positive feedback, negative feedback and randomness. The goal of such analysis is to recognize whether a stock is reverting to previous trading levels or if it is trending, and if so, on what time scale? Correctly analysing this aspect is absolutely necessary for making accurate market predictions.
One such application would be the “I Know First” market forecasting application. The goal of this application is to add new market trading data to the database daily, which could contain of up to 15 years of market data, so when it runs a learning and prediction cycle, it can create market predictions for the short and long term future. This application gets more accurate at predicting future market prices as the database size increases and the algorithm gets more accurate, so should provide more accurate and profitable indicators for buying / selling. Although this application may be difficult to use initially, the benefits for the future could be highly profitable and minimise the risk of losses incurred from Financial trading. ( I Know First )