search

GretaTimaite / UNBigDataHackathon2022

https://gretatimaite.github.io/campr/
0 stars 0 forks source link

sampling #5

Open GretaTimaite opened 1 year ago

GretaTimaite commented 1 year ago

A few ideas for sampling:

  1. random sample from all the countries
  2. 1 country per continent
  3. based on GDP
  4. based on where we're from :)
  5. clustering algorithm, such as kmeans, hierarchical...
    • what would you cluster based on?
  6. random sample per each category in GDP per capita as in World Bank data: https://data.worldbank.org/indicator/NY.GDP.PCAP.CD?view=map
GretaTimaite commented 1 year ago
  1. kmeans: GDP per capita and CO2 emissions, land temperature
GretaTimaite commented 1 year ago

@Kika0 is on fire!!!

GretaTimaite commented 1 year ago

done some means sampling, here's a result image

But then we kind of thought that maybe it makes sense to predict support for environmental protection to all countries...

EDIT: code can be found in sampling.R

GretaTimaite commented 1 year ago

Also I've realised that World Values Survey might not have data on EACH country, so I'll subset data based on this...