Mohamed AliApplied AI Engineer
PROJECT / 05Visual analytics · Geographic data

Mapping where new coders live

The large-city measure is a country-level survey percentage, not a label applied directly to the country or its total population.

Role
Geographic data integration and visualization
Context
Four-person JKU Visual Analytics project
Period
2022
INTERACTIVE EVIDENCE

Compare both maps.

Two country maps compare 2018 HDI with the percentage of surveyed new coders living in cities of more than one million people.

Countries with n ≥ 20
Human Development Index by countrySelect a country to compare its 2018 HDI with the percentage of surveyed new coders living in cities of more than one million people. Countries with fewer than 20 usable responses are hatched and carry no value.
HDI 0.450.97fewer than 20 responses — not shown
GUIDED STOPS
THE QUESTION

For each country, what percentage of surveyed new coders live in cities with more than one million people, and how does that compare with the country’s 2018 HDI?

HOW I APPROACHED IT
  1. 01

    Normalized inconsistent country names and generated standard country identifiers.

  2. 02

    Grouped usable responses by country and counted respondents whose city-population category was ‘more than 1 million’.

  3. 03

    Calculated that count as a percentage of usable country responses, then joined it with 2018 HDI and geographic boundaries.

  4. 04

    Displayed country aggregates only when at least 20 usable survey responses were available.

The call I made
Chose
Displayed a country only at 20 or more usable responses
Instead of
Mapping every country present in the survey
Because
A percentage from a handful of responses looks identical to one from thousands. The threshold stops the map implying precision the sample cannot support.
Survey28,397 responses
Public view91 country aggregates
Display thresholdn ≥ 20 per country
TOOLS & METHODS
  • Python
  • Pandas
  • GeoPandas
  • Survey aggregation
  • Choropleths