In kenarab/COVID19: COVID-19 data processing

knitr::opts_chunk$set(
  collapse = TRUE,
  dpi=200,
  comment = "#>",
  fig.path = "man/figures/README-africa-",
  out.width = "100%"
)

COVID19analytics

This package curate (downloads, clean, consolidate, smooth) data from Johns Hokpins for analysing international outbreak of COVID-19.

It includes several visualizations of the COVID-19 international outbreak.

Yanchang Zhao, COVID-19 Data Analysis with Tidyverse and Ggplot2 - China. RDataMining.com, 2020.

URL: http://www.rdatamining.com/docs/Coronavirus-data-analysis-china.pdf.

• COVID19DataProcessor generates curated series
• visualizations by Yanchang Zhao are included in ReportGenerator R6 object
• More visualizations included int ReportGeneratorEnhanced R6 object
• Visualizations ReportGeneratorDataComparison compares all countries counting epidemy day 0 when confirmed cases > n (i.e. n = 100).

Consideration

Data is still noisy because there are missing data from some regions in some days. We are working on in it.

Package

| Release | Usage | Development | |:--------|:------|:------------| | | | | | | | | ||||

How to get started (Development version)

Install the R package using the following commands on the R console:

# install.packages("devtools")
devtools::install_github("rOpenStats/COVID19analytics", build_opts = NULL)

How to use it

library(COVID19analytics) 
library(dplyr) 

data.processor <- COVID19DataProcessor$new(provider = "JohnsHopkingsUniversity", missing.values = "imputation")

#dummy <- data.processor$preprocess() is setupData + transform is the preprocess made by data provider
dummy <- data.processor$setupData()
dummy <- data.processor$transform()
# Curate is the process made by missing values method
dummy <- data.processor$curate()

current.date <- max(data.processor$getData()$date)

rg <- ReportGeneratorEnhanced$new(data.processor)
rc <- ReportGeneratorDataComparison$new(data.processor = data.processor)


top.countries <- data.processor$top.countries
international.countries <- unique(c(data.processor$top.countries,
                                    "China", "Japan", "Singapore", "Korea, South"))
africa.countries <- sort(data.processor$countries$getCountries(division = "continent", name = "Africa"))

# Top 10 daily cases confirmed increment
(data.processor$getData() %>%
  filter(date == current.date) %>%
  select(country, date, rate.inc.daily, confirmed.inc, confirmed, deaths, deaths.inc) %>%
  arrange(desc(confirmed.inc)) %>%
  filter(confirmed >=10))[1:10,]

# Top 10 daily deaths increment
(data.processor$getData() %>%
  filter(date == current.date) %>%
  select(country, date, rate.inc.daily, confirmed.inc, confirmed, deaths, deaths.inc) %>%
  arrange(desc(deaths.inc)))[1:10,]

rg$ggplotTopCountriesStackedBarDailyInc(included.countries = africa.countries,
                                                  countries.text = "Africa")
rc$ggplotComparisonExponentialGrowth(included.countries = africa.countries, min.cases = 20)

rg$ggplotCountriesLines(included.countries = africa.countries, countries.text = "Africa countries",
                        field = "confirmed.inc", log.scale = TRUE)
rc$ggplotComparisonExponentialGrowth(included.countries = africa.countries, 
                                     field = "deaths", y.label = "deaths", min.cases = 1)

rg$ggplotTopCountriesStackedBarDailyInc(top.countries)
rc$ggplotComparisonExponentialGrowth(included.countries = international.countries, 
                                               min.cases = 100)

rg$ggplotCountriesLines(field = "confirmed.inc", log.scale = TRUE)
rg$ggplotCountriesLines(field = "rate.inc.daily", log.scale = TRUE)

rg$ggplotTopCountriesPie()
rg$ggplotTopCountriesBarPlots()
rg$ggplotCountriesBarGraphs(selected.country = "Ethiopia")

kenarab/COVID19 documentation built on Oct. 1, 2021, 11:37 p.m.