COVID-19 Data Analysis & Geospatial Visualization Using Python

An empirical epidemiological and geospatial visualization investigation examining the worldwide trajectory and transmission dynamics of COVID-19 using official surveillance data from the Center for Systems Science and Engineering (CSSE) at Johns Hopkins University (JHU) spanning the complete 3-year global pandemic horizon (January 2020 through March 2023).

We fuse longitudinal infection and mortality metrics with national subjective well-being indicators from the United Nations World Happiness Report to analyze multivariate statistical correlations between virus transmission acceleration ($v_{\max}$) and socioeconomic life factors (GDP per capita, social support networks, healthy life expectancy, and civil liberties).

---

📑 Table of Contents

• Datasets
• Notebook 01: Epidemiological & Socioeconomic Study
  • 1. Longitudinal Epicenter Trajectories
  • 2. Transmission Velocity & Acceleration Peaks
  • 3. Distribution Normalization (Raw vs. Log-Transformed)
  • 4. Pearson Correlation Matrix Heatmap
  • 5. 4-Panel Socioeconomic OLS Regression Dashboard
  • 6. Multidimensional Prosperity vs. Velocity Matrix
• Notebook 02: Geospatial & Epidemiological Dashboard
  • 1. Executive Global KPI Metric Dashboard
  • 2. Comparative Caseload & Mortality Analytics
  • 3. Pandemic Severity Quadrant Classification
  • 4. Pareto Cumulative Mortality Concentration
  • 5. Population-Standardized Infection Intensity
  • 6. Interactive Geospatial Choropleth World Maps
  • 7. CFR Distribution & Diagnostic Ascertainment Bias
  • 8. Longitudinal Multi-Wave Timeline (2020 – 2023)
• Key Epidemiological Insights
• Environment Setup & Installation
• Connect with Me

---

📂 Datasets

1. COVID-19 Confirmed Cases Time-Series: Full historical daily confirmed cases across 289 jurisdictions and 201 sovereign nations (Jan 22, 2020 – Mar 9, 2023). data/covid19_confirmed_global.csv (Source: JHU CSSE)
2. COVID-19 Deaths Time-Series: Cumulative death records over 1,143 daily observation intervals. data/covid19_deaths_global.csv (Source: JHU CSSE)
3. COVID-19 Final Global Benchmark Snapshot: Comprehensive cross-sectional snapshot tracking Confirmed, Deaths, Case Fatality Ratio (CFR %), and Incident Rates across all global territories. data/covid19_daily_report_latest.csv (Source: JHU CSSE 03-09-2023)
4. World Happiness Report: National life factor indices scored across 156 sovereign nations covering GDP per capita, Social Support, Healthy Life Expectancy, Freedom to make life choices, Generosity, and Perceptions of Corruption. data/world_happiness_report_2019.csv

---

📓 01_COVID19_Data_Analysis.ipynb

Epidemiological Transmission Velocity Modeling & Socioeconomic Feature Fusion

1. Longitudinal Epicenter Trajectories

Tracks the multi-year cumulative confirmed caseload trajectories across the primary global epicenters (United States, India, Brazil, United Kingdom, Germany, and China) over the full 1,143-day surveillance horizon.

Figure 1.1: Cumulative confirmed COVID-19 trajectories across primary epicenters (Jan 2020 – Mar 2023), illustrating distinct infection milestones and exponential growth inflection points.

2. Transmission Velocity & Acceleration Peaks

Computes the first discrete derivative ($d(\text{Cases})/dt$) alongside centered 7-day moving average smoothing to eliminate day-of-week reporting artifacts and identify the exact all-time single-day maximum infection acceleration peak ($v_{\max}$).

Figure 1.2: Longitudinal daily infection velocity and 7-day smoothed moving average for the United States (Omicron peak: 1.03M cases/day) and India (Delta peak: 414k cases/day).

3. Distribution Normalization

Raw maximum daily infection rates exhibit extreme positive skewness ($>10^6$ in the US vs $<10^2$ in smaller territories). Applying a logarithmic transformation ($\ln(v_{\max})$) produces a bell-shaped Gaussian distribution suitable for parametric regression and correlation modeling.

Figure 1.3: Empirical density distributions (Histogram & KDE) of raw peak daily infection rate (severe positive skew) vs. log-transformed velocity ln(v_max).

4. Pearson Correlation Matrix Heatmap

Evaluates pairwise linear relationships between national socioeconomic happiness pillars and pandemic metrics, revealing strong positive associations between prosperity indices and recorded transmission velocity.

Figure 1.4: Annotated Pearson correlation matrix heatmap linking national socioeconomic happiness indicators to logarithmic transmission velocity and Case Fatality Rates.

5. 4-Panel Socioeconomic OLS Regression Dashboard

Four-panel Ordinary Least Squares (OLS) linear regression grid with 95% confidence intervals demonstrating robust positive correlations between national development indicators and peak transmission velocity:

• GDP per capita vs. $\ln(v_{\max})$: $r = +0.644$
• Social Support vs. $\ln(v_{\max})$: $r = +0.551$
• Healthy Life Expectancy vs. $\ln(v_{\max})$: $r = +0.640$
• Freedom to Make Life Choices vs. $\ln(v_{\max})$: $r = +0.470$

Figure 1.5: 4-Panel OLS regression grid with 95% confidence intervals and epicenter callouts showing the positive correlation paradox between economic development and recorded pandemic velocity.

6. Multidimensional Prosperity vs. Velocity Matrix

High-dimensional scatter analysis interlinking economic prosperity, healthy life expectancy, gross caseload volume, and peak transmission velocity across sovereign nations.

Figure 1.6: Multidimensional matrix mapping GDP Per Capita against ln(Peak Daily Cases), scaled by cumulative confirmed caseload and colored by Healthy Life Expectancy.

---

📓 02_COVID_19_Data_Visualization.ipynb

Global Benchmarks, Severity Quadrants, Pareto Mortality & Geospatial Mapping

1. Executive Global KPI Metric Dashboard

Summary of worldwide pandemic metrics benchmarked as of the final Johns Hopkins University CSSE surveillance snapshot (March 2023).

Figure 2.1: Executive KPI surveillance card panel summarizing aggregate global confirmed cases, cumulative casualties, global Case Fatality Rate, and tracked sovereign jurisdictions.

2. Comparative Caseload & Mortality Analytics

Comparative assessment of the Top 15 nations with the highest cumulative confirmed infections alongside a paired visualization comparing total mortality against Case Fatality Rates (CFR %).

Figure 2.2: Horizontal ranking of the Top 15 sovereign nations by cumulative confirmed caseload using a graduated mako color palette.

Figure 2.3: Paired comparative visualization of cumulative fatalities (left) alongside Case Fatality Rates (right), benchmarked against the 1.02% global average baseline.

3. Pandemic Severity Quadrant Classification

A bivariate log-log scatter plot categorizing nations into 4 risk quadrants based on median confirmed cases ($1.04 \times 10^5$) and median deaths ($1,385$).

Figure 2.4: Log-log pandemic severity quadrant classifying sovereign nations by transmission scale, mortality burden, and Case Fatality Rate intensity.

4. Pareto Cumulative Mortality Concentration

Empirical validation of the Pareto principle ($80/20$ rule) in global mortality. A dual-axis curve demonstrates that just 18 nations accounted for ~75% of all cumulative COVID-19 casualties worldwide.

Figure 2.5: Dual-axis Pareto distribution illustrating cumulative death counts by country and the cumulative percentage curve crossing the 75% threshold at nation 18.

5. Population-Standardized Infection Intensity

Contrasts gross caseload leaders with population-standardized Incident Rates (confirmed infections per 100,000 residents). High-testing and compact European nations (San Marino, Cyprus, Andorra, Austria, Portugal) exhibited extreme per-capita incidence (>60,000 per 100k) far exceeding raw volume leaders.

Figure 2.6: Horizontal lollipop chart comparing the Top 15 highest per-capita incident rate nations against their gross cumulative caseloads.

6. Interactive Geospatial Choropleth World Maps

Geospatial choropleth world maps tracking the worldwide distribution of cumulative infections and mortality severity.

Figure 2.7: Worldwide geospatial choropleth map illustrating cumulative confirmed COVID-19 infections on a logarithmic intensity scale.

Figure 2.8: Worldwide geospatial choropleth map illustrating sovereign Case Fatality Rates (CFR %) capped at 10% on a Natural Earth projection.

7. CFR Distribution & Diagnostic Ascertainment Bias

Examines global mortality variance and proves the testing ascertainment paradox: nations with low diagnostic testing rates experienced artificially inflated CFRs due to severe under-ascertainment of mild and asymptomatic infections.

Figure 2.9: Distribution density (KDE) and quartile boxplot showing the global Case Fatality Rate dispersion (global mean: 1.37%, median: 0.98%).

Figure 2.10: OLS regression showing testing ascertainment bias: low incident rate nations exhibited severely inflated CFRs, while high-testing jurisdictions converged to ~0.3–0.8%.

8. Longitudinal Multi-Wave Timeline (2020 – 2023)

Full 3-year epidemiological timeline tracking 14-day rolling average daily case counts across the major epicenters (United States, India, Brazil, France, Germany, Japan) with clear phase overlays:

• Phase 1: Wildtype Surge (Spring 2020)
• Phase 2: Delta Surge (Spring / Summer 2021)
• Phase 3: Omicron BA.1 / BA.5 Surges (Winter 2021 / 2022)

Figure 2.11: 3-Year multi-wave epidemiological trajectory tracking 14-day rolling average daily cases across epicenters, highlighting Wildtype, Delta, and Omicron wave surges.

---

🔬 Key Epidemiological Insights

1. The Positive Correlation Paradox ($r \approx +0.64$): Higher national GDP per capita and public longevity strongly correlate with higher peak daily infection acceleration. This does not imply wealth causes vulnerability; rather, it reflects diagnostic ascertainment capacity: high-GDP nations deployed mass RT-PCR screening, decentralized testing, and transparent digital surveillance, whereas developing nations faced test supply shortages, capturing only severe hospitalized cases.
2. Pareto Mortality Concentration: Global casualties were heavily clustered: 18 sovereign nations accounted for ~75% of cumulative global fatalities, underscoring the disproportionate impact on major transport hubs and older demographics.
3. Ascertainment Bias in Case Fatality Rates: Observed Case Fatality Rates varied from $0.11%$ (South Korea) to over $5%$ (e.g. Mexico, Peru, Yemen). As diagnostic testing density (Incident Rate) increases, recorded CFR drops sharply and stabilizes between $0.3%$ and $0.8%$.

---

🛠️ Environment Setup & Installation

Prerequisites

• Python 3.10+ (tested through Python 3.14)
• uv (recommended) or standard pip

1. Clone the Repository

git clone https://github.com/mohd-faizy/08P_COVID19_Data_Analysis_Using_Python.git
cd 08P_COVID19_Data_Analysis_Using_Python

2. Set Up Virtual Environment

Using uv (Lightning Fast):

uv venv
# On Windows:
.venv\Scripts\activate
# On Linux/macOS:
source .venv/bin/activate

uv add -r requirements.txt

Using standard pip:

python -m venv .venv
# On Windows:
.venv\Scripts\activate
# On Linux/macOS:
source .venv/bin/activate

pip install -r requirements.txt

3. Launch Jupyter Notebook / JupyterLab

jupyter lab
# or
jupyter notebook

---

🔗 Connect with Me