Citizen Science with Python

Citizen Science with Python

Given at PyConLT 2019, I talk on dirty air analysis in North Macedonia, monitoring relocated Orangutangs with drones and SciPy, improving child-birth with scikit-learn and improving political engagement with R and Python.

3d644406158b4d440111903db1f62622?s=128

ianozsvald

May 26, 2019
Tweet

Transcript

  1. Citizen Science with Python @IanOzsvald – ianozsvald.com Ian Ozsvald PyConLT

    2019 Keynote
  2.  Interim Chief Data Scientist  19+ years experience 

    How do we “do good” with data? Introductions By [ian]@ianozsvald[.com] Ian Ozsvald
  3.  Short stories on Citizen Science  Tips for how

    you could do the same  Short demo Goals today By [ian]@ianozsvald[.com] Ian Ozsvald
  4. Macedonian air quality By [ian]@ianozsvald[.com] Ian Ozsvald The “smelly fog”

    in Skopje Gorjan Jovanovski (when 21) Government open data showed “4* more pollution than Beijing and 20* EU limits” www.theair.app Photo: https://www.facebook.com/ilijoski
  5.  Initially a JSON dump  1 mil. People in

    1 mo.  Visualisation in Parliament  Challenged by a Minister  Drove Government policy Political debate and change By [ian]@ianozsvald[.com] Ian Ozsvald
  6.  Get a public dataset – most people can’t but

    you can  Load it, investigate  Make some graphs with matplotlib  Can you tell a story? How to start? By [ian]@ianozsvald[.com] Ian Ozsvald
  7. Updating outdated medical results By [ian]@ianozsvald[.com] Ian Ozsvald

  8. Updating outdated medical results By [ian]@ianozsvald[.com] Ian Ozsvald  Friedman

    1955  Stages of labour by cervix dilation  Different drugs, ages, technologies  Significant medical decisions based on the result
  9. Updating outdated medical results By [ian]@ianozsvald[.com] Ian Ozsvald

  10. Actionable results By [ian]@ianozsvald[.com] Ian Ozsvald First or ‘many’ births?

    Weight (before delivery) Age
  11.  Check for out-dated assumptions  Gather data & visualise

     Produce simple models (if/then models are great!)  Build up iteratively as you gain trust How to start? By [ian]@ianozsvald[.com] Ian Ozsvald
  12.  Dirk Gorissen – track 6 Orangutangs in 2000 km^2

    of Bornean jungle  Radio pings, drones and signal processing Where’s Bunga the Orangutang By [ian]@ianozsvald[.com] Ian Ozsvald
  13.  Sw Defined Radio, Python processing  Automated flight paths

     SEE VIDEO  Success on second round! Where’s Bunga the Orangutang By [ian]@ianozsvald[.com] Ian Ozsvald https://dirkgorissen.com/2019/01/06/wheres-pinoh-tracking-orangutans-with-drones-and-gnu-radio/
  14.  Automate a manual process  Note – drones are

    complex, don’t start here!  Collect data, analyse, visualise signal, make decisions How to start? By [ian]@ianozsvald[.com] Ian Ozsvald
  15.  Hackathon  Can we get more people registered to

    vote?  Using only open data and adwords  Thanks Alex, Emiliano, Harry, John, Laurens, Nick Improving political engagement By [ian]@ianozsvald[.com] Ian Ozsvald
  16. Improving political engagement By [ian]@ianozsvald[.com] Ian Ozsvald

  17. Improving political engagement By [ian]@ianozsvald[.com] Ian Ozsvald

  18.  Jupyter Notebooks and open data  Can we challenge

    a Guardian news article and fix it?  Thanks to Tom Forth @thomasforth Short demo By [ian]@ianozsvald[.com] Ian Ozsvald
  19.  Tell useful data stories  Where’s PyData Vilnius? Start

    it?  Thank your organisers & speakers!  Send me a postcard if I taught you something? Summary By [ian]@ianozsvald[.com] Ian Ozsvald
  20. “Monitoring Personal Air Quality” By [ian]@ianozsvald[.com] Ian Ozsvald Github.com/OxygenLithium PyLondinium

    2018 talk by Robin & Oliver