I can honestly say that today wasn’t too bad considering…

Our mission was to access the Food Standards Authority’s data regarding hygiene ratings, fortunately they have an API, albeit a poorly documented one, and to make it a little more challenging the documentation was set up for XML scraping, something we had only a little experience in.

Anyway, due to some technical hiccups and the daunting realisation that the dataset was huge (no seriously, at 650GB we decided to call it quits and focus our efforts on our great city of London) we didn’t have the data ready to play with until after lunch. However, once that was useable I started to dig in.

Over lunch (a lovely teriyaki chicken) I took my notepad with me (#pro tip) and started to scribble down ideas and possible findings that came to my mind as I ate and watched the world go by (the inclusion of house price data came from a Foxton’s mini driving by….).

I thought I would look at hygiene ratings by local authority area and see if I could loop in some data about house prices to see if they affected each other – spoiler they didn’t.

I also wanted to include a map showing areas and businesses to avoid but unfortunately as the geo-coding only recognised the first part of the postal code or local authority (with the help of Craig Bloodworth’s tools) this couldn’t be made granular enough and just looked like the previous map, so that didn’t make the final cut.

Lastly I was recently advised by the great @TableauTim to always try and include some amount of interactivity in my work, so I built a simple view that allows you to enter your post code and it lists all the nearby businesses, their address, and their associated food hygiene rating.

As I made the postcode finder it made me think of the endless apps on my phone that want to track my location, and realised that I hadn’t really made a decent mobile designed viz yet, so I built my dashboard to fit on a teeny tiny mobile phone screen!

Anyway, if you want to see the result, it’s here (open it on your phone for the best results!):