ECON30006: Data Science

by Athirah Ridzal

Week 1 - Getting Started

These charts were taken from Rapid Charts website. The charts plots the COVID-19 cases by UK regions, UK labour productivity and the increasing trend of aging populations.

Week 2 - Hosting Data

The charts shows the number of first dose vaccinations and death dates from COVID-19 in three UK cities.

Week 3 - Editing Data

The first chart plots the predicted global population added into the csv file. The second shows the average daily steps I've taken from April to October, data manually written in JSON format.

Week 4 - API Driven Charts

These charts are hosted by APIs. The first chart is from OECD that plots the gender wage gap for South Korea and total OECD countries. The OECD API has an uncommon JSON structure, thus an XML converter was used. The second chart is from WHO, showing the share of alcohol related crashes for five countries across 20 years. The WHO API can't be access directly through Vega, hence a CORS reverse proxy was used.

Week 5 - Raw Data from Github

These charts are US based data that shows the increasing trend of the wealth share of the top 1%, total expenditure on educational books, the savings rate and the slight decreasing growth of debt service for households. The data for these were batch downloaded using a for loop using Python in Colab.

Week 7 - Simple Scraper

This is a chart of the 10 largest colonial empires in human history, data scraped from World Atlas.

Week 8 - Data Story

At the Talking Economics event, one of the panelist, Jagjit Chadha, the director of the NIESR stated that the UK had one of the largest drop of output and is experiencing much slower recovery amongst the advanced economies. Amongst the G7 countries, UK had the biggest drop in 2020 at -9.79%. Nonetheless, according to the UN, for 2021 Japan and Germany will on average have a lower expected growth than the UK. By 2022 and the upcoming years, the G7 economies will follow similar trends, where the UK is expected to grow more than most of the other countries.

Week 9 - Advanced Analytics

This first chart regresses the public healthcare expenditure per capita on the dependency ratio of the older population using a polynomial at the 6th degree. The polynomial regression fits the observations best than the common linear regression and any lower degree polynomial, with an R² of 0.60. The second chart is the linear regression of the dataset with an R² of 0.52. The polynomial regression takes account the variation observations across the dataset, hence better describing the data. In general, both charts describe the increase of public spending as the dependency ratio of the older population increases. The data cleaning where an outlier with an dependency ratio of 45 was removes and the fitted lines were created using Python in Colab.

Week 10 - Interactive

This is an interactive chart showing the percentage of youth that are not in employment, education or training for 4 OECD countries and the OECD average. The range of years can be changed using the sliders and the value of each data point can be observed by hovering over the lines.