Return to site

Making a hash out of knitting with data

· Data,Standards,Craft,Knitting,Open data

Over the summer of 2018 I spotted on social media that someone had created a scarf that recorded the train delays that she’d encountered on her commute. I loved this. Data is so powerful when it’s tangible - it sparks conversations and generates new ideas.

It chimed a lot for me with Stef Posavec’s work that I’d seen in the Big Bang Data exhibition in London a few years earlier. It was also a real treat to watch the progress on Open Data Manchester’s data loom, where they had woven a piece of cloth using the initials of people who visited their tent at the BlueDot festival in 2019. And the two pieces of cloth woven from the election results from 2017 and 2019. So with all this inspiration and my tendency towards a bit of nature and craft downtime, it was inevitable that I would eventually get to work on my own data knitting project.

Julian from Open Data Manchester holding a striped woven cloth depicting election data. Julian is pointing to towards the data for Chi Onwurah's  constituency. Chi  stands smiling in the background.

Julian Tait, CEO of Open Data Manchester CIC, with Chi Onwurah, MP for Newcastle Central, and the election data cloth

I’ve been doing the RSPB Big Garden Birdwatch for several years - I love a bit of citizen science. Each year, I select a one hour slot during the specified weekend in January, watch the birds and count the maximum number of each species that lands at any one time, then upload the results to the RSPB site.

Last year, I didn’t stop there. I used a cheat sheet from Open Data Manchester’s loom projects that converts the alphabet to ITA2 Baudot-Murray code. There are 5 rows of knitting needed for each letter of the alphabet, so spelling out bird names wasn’t really an option unless I wanted to produce a scarf worthy of Tom Baker’s Dr Who.

I needed a simple code list. After a fair bit of digging I discovered The British List, which had an option of 2 or 5 letter codes for birds recorded in Britain. 10 rows of knitting per bird seemed reasonable, so I set about translating my bird names to bird codes, so I could get to the Baudot-Murray codes. That would require two colours for the binary representation. I used Rowan Hemp Tweed in Duck Egg for 0s and Paprika for 1s. So the letter A was:

  • Duck Egg
  • Duck Egg
  • Duck Egg
  • Paprika
  • Paprika

Some of the 2 letter codes were actually a single letter (standards, eh?), so I thought I might get into a tangle if I couldn’t easily see where the spaces were between bird types. I introduced a third colour (Rowan Hemp Tweed in Teal) to help me navigate the pattern visually. So for SPACE in the Baudot-Murray Code I used:

  • Duck Egg
  • Duck Egg
  • Teal
  • Duck Egg
  • Duck Egg

If I saw only one of a species of bird, I knitted its code once, if I saw 3 of them, I repeated the code three times. I added SPACE only between different types of bird, so I could get a sense of which were the most prevalent, however some of the codes being a single letter rather than 2 makes it less visually reliable.

What I found fascinating about this project was how much I learned about the whole experience of implementing a simple standard using code, and the decisions I had to make along the way. It came together in a way that worked, but someone else repeating the exercise starting with the Baudot-Murray codes, the Bird List, and the data from my bird count, would have ended up with a very different outcome. If they’d used their own data, we definitely wouldn’t have been able to lay our scarves side by side and draw any comparisons about the birds that had visited our gardens.

It’s ever likely that there are so many interoperability problems in the software and data world. To reach a shared outcome, co-operation on standards really does matter.

All Posts
×

Almost done…

We just sent you an email. Please click the link in the email to confirm your subscription!

OKSubscriptions powered by Strikingly