Group15 proposal

Proposal

Poster

Application

Research Paper

Back to Main

Overview

Since December 2019, new cases of COVID-19 have been discovered in Wuhan, it has spread to all over China and other countries. As of 2019-03-01, a total of 79,971 cases have been diagnosed and a total of 2,873 deaths have been confirmed in China. The situation in the rest of the world is not optimistic, 7,277 cases have been confirmed and the total 18 deaths occurred in other countries. Its widespread and rapid speed is alarming. The 2019 nCov has caused a certain impact on the economy of China and even countries around the world, even has been taken seriously by WHO. In order to study trends in the number of diagnoses, the number of suspected patients, and deaths over time for different areas, we use the 2019 nCov data for further analysis and visualization.

Project Objectives

We plan to use the 2019 nCov data to:

1)Analyze trends in cumulative cure rates and mortality across the country and provinces.

2)Analyze trends in new diagnoses and mortality rates across the country and provinces.

3)Using multiple measurement methods to display epidemic data maps-based on nuclear probability density, based on clustering, etc.

Data Source

Since the outbreak of the new coronavirus pneumonia, many agencies/portals (Clove Garden, Tencent, Sina, ...) have provided pages for "real-time tracking of the epidemic", which mainly displays the number and area of suspected, confirmed, cured, and dead cases distributed, the data used in this project is collected from the "Clove Garden "website.

The data is from the website below. It includes Province name(Chinese and English), City name(Chinese and English), zip code, the province confirmed count, province suspected count, province cured count, province, death count, the city confirmed count, city suspected count, city cured count, city dead count and update time.

https://ncov.dxy.cn/ncovh5/view/pneumonia

Methodology

(1)Data Exploration and Visualization: We will make full use of the dataset to visualize the data in different ways. Maps are the most intuitive way to show the spread and control of the epidemic. Line charts are the best way to show data trends and forecast data.

(2)Clustering: Cluster the city and cases according to the daily addition rate, cured rate, and death rate.

(3)Time Series Analytics: Based on the current data to predict the future trend for every single city.

Tools&Packages

Tools: Tableau; R-shiny

Packages: magick, dplyr, ggplot2, shinydashboard, forcats, mapproj, etc.