Why forecast sales?
Humans have the magical ability to plan for future events, for future gain. It’s not quite a uniquely human trait. Because apparently ravens can match a 4-year-old.
An abundance of data, and some very nice R packages, make our ability to plan all the more powerful.
A couple of months ago we looked at sales from an historical perspective in Digital Marketplace. Six months later. In this post, we’ll use the sales data to March 31st to model a time-series forecast for the next two years. The techniques apply to any time series with characteristics of trend, seasonality or longer-term cycles. Continue reading “But can ravens forecast?”
With tensions heightened recently at the United Nations, one might wonder whether we’ve drawn closer, or farther apart, over the decades since the UN was established in 1945.
We’ll see if we can garner a clue by performing cluster analysis on the General Assembly voting of five of the founding members. We’ll focus on the five permanent members of the Security Council. Then later on we can look at whether Security Council vetoes corroborate our findings. Continue reading “An East-West less divided?”
Revisiting an old post
Last September I wrote a post entitled Is the Government realising its ambition for SMEs on G-Cloud? Six months on, I wanted to revisit and update this article, fold in a second Digital Marketplace framework, and share the R code here. Revisiting an old post also provides an opportunity to see if one can simplify and improve older code. Continue reading “Digital Marketplace. Six months later.”
Supervised machine learning
In the “cluster of six”, we used unsupervised machine learning, to reveal hidden structure in unlabelled data, and analyse the voting patterns of Labour Members of Parliament. In this blog post, we’ll use supervised machine learning to see how well we can predict crime in London. Perhaps not specific crimes. But we can use recorded crime summary data at London borough-level (non-personal aggregated data licensed under the Open Government Licence), with some degree of accuracy, to predict crime counts.
Along the way, we’ll see the pay-off from an exploration of multiple models.
Continue reading “Criminal goings-on in a random forest”
Every story needs a good plot
One could think of data science as “art, grounded in facts”. It tells a story through visualisation. Both story and visualisation rely on a good plot. And an abundance of those has evolved over time. Many have their own dedicated Wikipedia page!
Which generate the most interest? How is the interest in each trending over time? Try this app to find out. Continue reading “The plots thicken”
Unsupervised machine learning
Hansard reports what’s said in the UK Parliament, sets out details of divisions, and records decisions taken during a sitting. The hansard R package provides functions to import its data.
Using the Hansard API (Application Programming Interface), we’ll apply unsupervised machine learning to analyze the voting patterns of 219 Labour Members of Parliament (MPs). We’ll consider all divisions (results of the votes) in the UK House of Commons since the 2017 general election. Continue reading “The “cluster of six””
Responding to a weak property market
In December I looked at how recent events have shaped the property market in London SW10. If short-distance moves are off the table in the current climate, how are property owners responding? When sales are weak, are planning applications in the ascendency? I applied data science techniques to Royal Borough of Kensington and Chelsea (RBKC) planning data to find out.
Continue reading “SW10 digs deep”
Experimentation with geospatial mapping
Recently I experimented with geospatial mapping techniques in R. I looked at both static and interactive maps. Embedding the media into a WordPress blog would be simple enough with a static map. The latter would require (for me) a new technique to retain the interactivity inside a blog post.
My web-site visitor log, combined with longitude and latitude data from MaxMind’s GeoLite2, offered a basis for analysis. Although less precise than the GeoIP2 database, this would be more than adequate for my purpose of getting to country and city level. I settled on the Leaflet package for visualisation given the interactivity and pleasing choice of aesthetics.
The results however were a little puzzling.
Continue reading “Surprising stories hide in seemingly mundane data”
The anatomy of SW10
Analyses of house sales often focus on the wider UK market. In this blog, we’ll take a deep dive into one of London’s more-than 100 postcode districts. We’ll draw on 10,000 property transactions to see how key events have shaped the market. The object of our focus will be SW10 which forms part of the Royal Borough of Kensington and Chelsea.
We’ll start with the anatomy of SW10 per the chart below. Over 80% of property transactions were for leasehold flats / maisonettes. In contrast, detached freehold properties are a prized scarcity: Only 40 of the circa 10k transactions, over the past 20 years, were for detached properties. Continue reading “House sales in London SW10 take a few punches”
Why take a deeper look at G-Cloud categories?
The last blog – “The key to unlocking services on G-Cloud” – touched briefly upon their overlap. And as the concept of G-Cloud categories was newly introduced in the current iteration (G9), it may be worth taking a deeper look at their impact in advance of the next.
So, in this blog, I want to explore the extent and effects of category overlap. And let’s see what insights may be drawn. For example, are some categories of less value than others? Could some suppliers gain an advantage? Perhaps by aligning each service to many categories so buyers find them irrespective of their carefully crafted search criteria?
Continue reading “Do G-Cloud categories need a tweak?”