Below are libraries I used in this project,
pip install pandas
pip install numpy
pip install matplotlib.pyplot
pip install seaborn
pip install sklearn
Covid-19-Vulnerability-Analysis.ipynb: Covid-19 Vulnerability Analysis
data/Economic_Indicator.xlsx: Economic Indicator dataset for G20 countries
data/population_2020.csv: 2020 Population dataset at country level
data/COVID-19 Cases.csv: Johns Hopkins' county-level COVID-19 case and death data, paired with population and rates per 100,000
This project is focused to evaluate the current potential risk and opportunities at the country level (G20) based on covid-19 situation and economic indicators. In 2020, coronavirus heavily impacted all countries globally across industries, government did at all costs to control the situation, support the economy, and helped business and people through different programs. No one knows what would happen in each country and it might be the watershed for the next decade, which is the reason that I feel motivated and want to find it out that which countries are in good status with a solid economic foundation and which countries are at the edge of risk.
Currently, countries across globe at different stages fighting with coronavirus, which reflected how efficient the government handled the situation from the public health perspective. The economic indicators on the other hand reflected how strong the economy is as well as the effectiveness of the government financial programs. Leveraging covid-19 and economic info will help us to understand,
From public health perspective, which country will get recovered soon that all social and economic activities could go back on the regular track? Under the impact of covid-19 and flood of money injecting from government, which country's economy is strong enough to get recovered? By answering those two questions, we could get a full understanding to evaluate the countries' vulnerablities under this severe pandemic situation. Also, which countires are similar across economic measurements and covid-19 measurements will be another good insights. Hopefully through this analysis, we could uncover potential opportunites and make sure we are able to split the risks into different baskets.
By evaluating the pandemic situation and economic indicators for G20 countries, which countries are in good status with a solid economic foundation? and which countries are at the edge of risk? How similar the countries are based on the evaluation?
In order to help answering the questions related to Covid-19 and economics, I collected the info from Johns Hopkins COVID-19 Case Tracker and Trading Economics.
-
COVID-19 Cases.csv — containing daily confirmed cases, increases of a confirmed case, death case, and increases of death at the state level across countries.
-
Economic_Indicator.xlsx — containing economic indicators from actual to 2021 for countries in G20.
-
Population Dataset — containing population, change, urban population, world share, fertility rate, and middle age at the country level.
- For the COVID dataset, here are some initial processing steps,
a. Check the NULL values
b. Update the field type
c. Aggregate the stats at the country level
d. Transpose the data based on case type
e. Check the null value and drop the columns which are not needed
f. Pick the countries on the G20 list and update the name based on the economic dataset
- For the economic indicator dataset, here are some initial processing steps,
a. Update the measurement name to be consistent across countries
b. Select the measurements which are covered all G20 countries
c. Update the field type
d. Restructure the dataset to have each individual measure listed on the columns
e. Convert Balance of Trade and Current Account to the same currency unit (USD Million)
The evaluation will include the rank of involved metrics, which represented below.
A. Covid-19:
-
Mean number of confirmed and death case
-
The current number of confirmed and death case out of the total population
-
Avg Growth rate of confirmed and death case
-
Avg Death rate (death case/confirmed case)
B. Economic Indicator:
-
Currency
-
Stock Market (Points)
-
GDP Growth Rate (%)
-
Unemployment Rate (%)
-
Inflation Rate (%)
-
Interest Rate (%)
-
Current Account to GDP (%)
-
Government Debt to GDP (%)
-
Government Budget (% of GDP)
Corporate Tax Rate (%)
- Personal Income Tax Rate (%)
If we assume that the measurements have the same % impact, we could see the result below,
The rank across COVID-19 measurements represents the severity level of the pandemic based on the mean number of cases, growth rate, death rate, and a current number of cases. If the value is higher, the situation is much worse. From the visual above, in general, the pandemic situation in a couple of countries within Asia are in the recovery stage compared to the situation in Europe, North and South America.
The rank across economic measurements represents the solid foundation of each country. Except that the GDP Growth rate is higher, the economy is stronger, all the other indicators show a stronger foundation if the value is lower. In this final evaluation, if the value is higher, the situation is much worse. From the visual below, in general Switzerland, Russia, Singapore, Netherlands, Saudi Arabia have a stronger economic foundation.
From the final evaluation considering Covid-19 and economic foundation, Singapore, Russia, Switzerland are the top three countries with a stronger foundation and less server pandemic situation. On the other hand, Argentina, Spain, Brazil are the countries that are more vulnerable within G20 countries.
In addition to the rank of vulnerability across G20, what countries are similar within the same group regarding covid-19 situation and economic foundation?
Within the analysis, Kmeans, GaussianMixture, Affinity Propagation model are tested and the KMeans model with a higher score (silhouette) which is 0.1.
From the result, we could see,
-
Argentina is independent as one cluster, and the current status is very fragile based on the econ evaluation, which we should pay attention to and watch out.
-
Russia, Saudi Arabia, Singapore are clustered in one group, and in general, the COVID-19 situation is controlled well and the econ foundation is pretty strong. We could look for potential opportunities within.
-
Australia, South Korea, China, Germany are clustered in another group, and in general, the spread of COVID-19 is controlled and the econ foundation is comparatively strong. We could look for potential opportunities within.
-
United States, Switzerland, Netherlands, United Kingdom, Indonesia have a strong econ foundation but the COVID-19 situation is not controlled well.
-
India, South Africa, Canada, Brazil, Turkey, Mexico have some uncertainty based on COVID-19 and economic situation. France, Spain, Italy have a relatively worse pandemic situation.
In order to evaluate the clustering result, silhouette score is used. Based on its explaination, the best value is 1 and the worst value is -1. Values near 0 indicate overlapping clusters. Negative values generally indicate that a sample has been assigned to the wrong cluster, as a different cluster is more similar.
By evaluating the silhouette score, I will use the result from kmeans clustering and there is potential to improve to the clustering performance.
- kmeans_silhouette: 0.1
- gaussian_silhouette: 0.08
- affinity_silhouette:-0.07
From the result, we could see,
Argentina is independent as one cluster, and the current status is very fragile based on the econ evaluation, which we should pay attention to and watch out.
Russia, Saudi Arabia, Singapore are clustered in one group, and in general, the COVID-19 situation is controlled well and the econ foundation is pretty strong. We could look for potential opportunities within.
Australia, South Korea, China, Germany are clustered in another group, and in general, the spread of COVID-19 is controlled and the econ foundation is comparatively strong. We could look for potential opportunities within.
United States, Switzerland, Netherlands, United Kingdom, Indonesia have a strong econ foundation but there are some uncertainties if the COVID-19 situation will be controlled well or not.
India, South Africa, Canada, Brazil, Turkey, Mexico have some uncertainty based on COVID-19 and economic situation. France, Spain, Italy have a relatively worse pandemic situation.