/DI-test

Primary LanguagePython

Development Initiatives Test

How to:

  • clone this repo
  • create and load virtual environment in local directory
  • install requirements:
pip install -r requirements.txt

python run.py

Enjoy!

Task

Using Python, write functionality that takes any ISO country code as input, and can return for the given country: * A list of Global Adaptation Index scores for all available years * The average of all Global Adaptation Index scores for all available years

There's no need to return the country text name (e.g. Andorra if the input country code was AD)

Global Adaptation Index CSV

Technologies

  • Python (with csv module)
  • Pytest

Comments on Task

Time taken: ~9 hours

Decisions:

  • Created a method to loadcsv file so data can be updated in future if needed
  • Extracted data extraction from csv to a method that other methods depend on to make code more DRY and keep to SRP
  • Removed returnDataOptions method as task specifies that data will feed into external interface so a command line UI isn't needed
  • Added range and standard deviation because why not? But really because they gave a good indication of how the data has changed over time in relation to the mean value.
  • Didn't add mode or median because they didn't seem useful for this data set as scores didn't repeat and the middle score would have given more of a time snapshot than any useful data on score trends
  • Didn't want to finish with no working code so experimented with refactoring in separate files but ran out of time for writing new tests

Challenges:

  • Getting used to the csv module
  • Incomplete Python knowledge
  • Extracting the data in a useful way
  • Remembering how to do averages... it's been a while.
  • Problems with zip on refactoring(can only use list on zip once per dataset generation)
  • Not enough time to write tests for refactored code

Refactoring in progress

  • Separated responsibilities in original code to CountryData and Averages classes
  • All methods tested
  • After feedback made it so users can continue to select country info via ISO code until they choose to quit program
  • Finally got rid of the need for a CSV file. Though the climate-vulnerability.csv URL is hardcoded into the run file currently, purely to avoid errors from URLs with invalid formats being passed to the method
  • Looking to add some kind of URL data validation to allow URLs to be passed in as long as they have the correct format in the future
  • Need to add error handling for when users enter an ISO code that doesn't have any associated values