This is a short piece on a project I’m developing at NHS England, using Python to analyse patients’ inpatient care and determining where a Clinical Commission Group’s (CCG) patients are being treated significantly differently to those patients in its peer group CCGs (similar 10 CCGs as determined by NHS Rightcare Methodologies).
Category: General
Starting Back at Work
At the start of the month, I started back with my old team at NHS England working as (you may have already guessed it isn’t clinically based) a data analyst.
Florida Ticket Price Comparisons
I’m going on holiday soon! I’m headed to Florida with my parents and being the cost savvy Brits that we are we’re always on the hunt for a bargain. We wanted to go do some of the Theme Parks (we settled on Disney and/or Universal Studios) and went about finding the cheapest ticket rates we could. During the internet hunt my Mum kept using the Calculator on her PC to do things like exchange rate conversions and to calculate price differences – but, being the Excel monkey that I am I proposed a much better solution: let’s whack it all in Excel and let me make some pretty charts!
Zipf’s Law
After watching a long session of ‘VSauce’ videos (great brain food videos, albeit very addictive!), I came across this video discussing ‘Zipf’s Law’. Zipf’s law states that in any corpus of natural language, the frequency of any word is inversely proportional to its rank in the frequency table. This Zipfian distribution applies to many different types of data studied across a variety of fields (the video discusses a large variety of these instances). Zipfian distributions also follow the ‘Pareto Principle’, the 80-20 rule. 80% of the words used in any corpus are only 20% of the unique words used.
After being utterly captivated by this phenomenon, I decided it would be fun to check whether my blog, being a corpus of natural language, followed Zipf’s law.
Tinkering with Bullet Charts
After reading some of my other favourite data visualisation blogs (Storytelling with Data to be precise) I came across a new type of visualisation I’d not seen before – the Bullet Chart.
Once I saw Bill Dean combine the two-series bar chart into a bullet chart, I couldn’t help but think of my last blog post and the set of two-series charts I produced comparing British and Non British populations in London. Curious, I went about converting the charts I had made into this new format, and I’m very pleased with the results:
