Changing the world with data

Jumping Rivers

Dr Clarissa Barratt & Dr Rhian Davies

Part 1: Hello 👋

Clarissa

Helloo, I’m Clarissa! I am the Ambassador for Data Science at Jumping Rivers, which just means I talk a lot…

I started out doing a degree in Astrophysics, worked in cyber security for a bit, moved to Newcastle to do a PhD in Quantum Physics before starting to work for Jumping Rivers. I now work from home in Austria!

A cartoon of Clarissa

Clarissa

Rhian

  • I love helping people understand their data

  • I enjoy dabbling at lots of different things

  • I near Newcastle by the sea 🌊

A cartoon of Rhian

Rhian

An infographic showing various statistics about 2022 activity

🚣 Jumping Rivers

  • Small company in Newcastle
  • We work with a huge variety of domains including:

    • Farming

    • Railway

    • Energy

    • Covid-19

  • Provide data science and engineering services and training

A cartoon of a robot with a clipboard.

📔 The Plan

  • What is a Data Scientist?

  • Being a data detective with problem solving skills

  • Unleash the inner artist

🔍 What is consultancy?

  • General problem solving

  • Making peoples lives easier

  • Exists across all

A cartoon of a robot with a clipboard.

🦄 What is a data scientist?

A collection of cartoons of Jumping Rivers employees. Top: Clarissa, Myles, Keith, Russ. Bottom: Nicola, Jack, Rhian, Mandy

📝 Your turn

  • Introduce yourself with a visualisation
08:00

👋 You!

Part 2: Playing with Data

🍔 What’s your favourite food?

  • We asked you two questions:

  • What’s your favourite fast food restaurant?

  • What’s your favourite snack?

👩‍💻 Explore your data

👩‍💻 Data Quality issues

  • Biased sample (14-19 year olds)

  • Sentences rather than one-word

  • Spaces and capitalisation (e.g “Pringles” / ” pringles”)

  • Is there missing data?

  • Sometimes we have different names for the same things

  • Can we “validate” the data?

🔍 What’s the question?

  • What do we want to know?
  • What else do we need to know to answer the question?

📝 Your turn

  • What questions could you ask the data?

Try and be creative

02:00

👀 Plot examples

Wordcloud of common words in fast food answers. Mcdonalds, Subway, Greggs and KFC are largest.

Wordcloud of common words in snack answers. Crisps and chocolate are largest.

🔋Power Up your data with transformation

  • Creating data “features” by combining with other data set
  • Linking data sets together
  • Why do we transform?
  • Data we don’t have that we want to pull in

🔋What data-sets might we want to add

  • Nutritional data
  • Classifying sweet/savoury, chain/independent
  • Location data - we could make maps!
  • Average price of the fast food restaurant?
  • Advertising budget of the restaurant/chain?
  • Classifying to create small categories
  • E.g. Pizza places with different names
  • Savoury / sweet

👀 Examples

👀 Examples

👀 Examples

🍔 What type of fast food?

🧲 Let’s join

🧲 Left Join

🧲 Full Join

🧲 Inner Join

Part 3 Visualisation

🔍 What types of graphs can we make?

  • Sentence
  • Table
  • Bar chart
  • Scatter plot
  • Can be interactive

🔍 What makes a good viz

  • Tells a story
  • Clear and concise
  • Truthful!
  • Engaging title (it doesn’t just have to say “A graph to show …”)

👀 Viz examples

Map of the London underground

👀 Viz examples

Football match with pundits commentary highlighted.

👀 Viz examples

Plot of 20 and 40 year cycles in the stock market pointing to a low in 2022.

👀 Viz examples

Plot of many coloured lined interacting on a black background. There are no axes and there is no text.

🐧 Meet your clients

Poorly edited image of Rhian and Clarissa cartoons with bobble hats on, in front of an icy mountain, with three cartoon penguins between them.

📝 Your turn

You are the data science consultant. We want to learn something about our penguins

  • What do you want to ask the client?
  • What can you learn from the data?
  • What does the client want to learn?
03:00

🐧 Penguins

📝 Your turn

Play around with the shiny app to build a scatter for our penguins.

jr-ctw.jmpr.io/scatter

10:00

👀 What makes a great viz?

  • What do you like about this graphic?
  • What makes a good title?
  • Do you need legends and labels?
  • Does it tell a story?

{fig-alt=“A cartoon of a woman with acne. The spots on her face are real data points” fig-align=“right” width=“350” caption = “https://monachalabi.com/}

🎨 Bob Ross - Libby Daniells

🎨 Bob Ross - Jared Wilber

🎨 Bob Ross - Ryan Hart

🎨 Bob Ross - Nicola Rennie

📝 Your turn

  • Work together in the Jam board to design a great graphic
  • It doesn’t have to be accurate - we’re just looking for a sketch/design
20:00

The Wrap up

Something we made earlier

Thanks! 👋

Jumping Rivers