This research investigates the impact of natural events on public and economic health. The data used is the National Weather Service Storm Data collection over 61 years (from 1950 untill 2011) which contains 902297 observations and 37 variables. The purpose of this report is to show which weather related event types have the most damaging impact. The first graph shows the top eight event types which lead to the most direct peronal injuries and the second graph shows the top eight event types which lead to most economic damage. Both graphs clearly show the Tornado event type as a leading cause in both personal injuries and economic damage, with other wind or water related (e.g. flooding, flash-flooding) event types a not so close second. This research suggests taking a further look into preventing wind and water related disasters as a measure with the highest ROI.

Data Processing

library(ggplot2)
# working directory set in console
## Download the dataset into a new directory
fileurl <- "https://d396qusza40orc.cloudfront.net/repdata%2Fdata%2FStormData.csv.bz2"
datadir <- "./data"
zipfile <- "projectdataset.bz2"
zipfilefullpath <- paste(datadir, "/", zipfile, sep="")

# create data dir if it doesn't exist yet
if(!file.exists(datadir)){dir.create(datadir)}

# download data file if it wasn't downloaded before
if(!file.exists(zipfilefullpath)){
  download.file(fileurl, destfile = zipfilefullpath)
}

d <- read.csv(zipfilefullpath)

Results

Across the United States, which types of events (as indicated in the EVTYPE variable) are most harmful with respect to population health?

require(reshape2)

#transform data
d.aggr <- aggregate(cbind(FATALITIES, INJURIES) ~ EVTYPE, data = d, FUN = sum)
d.order <- d.aggr[order(d.aggr$INJURIES, decreasing = TRUE),]
d.head <- head(d.order, 8)
d.melt <- melt(d.head, id.var="EVTYPE")

#create plot
plot1 <- ggplot(d.melt, aes(x = EVTYPE, y = value, fill = variable)) +
  geom_bar(stat="identity") + 
  theme(axis.text.x = element_text(angle=45, hjust=1, vjust=.95)) +
  ylab("Number of people affected") + xlab("Event Type")
plot1

#show data used in a table
head(d.head, 8)

Across the United States, which types of events have the greatest economic consequences?

#transform data
eco <- aggregate(cbind(PROPDMG, CROPDMG) ~ EVTYPE, data = d, FUN = sum)
eco.ordered <- eco[order(eco$PROPDMG, eco$CROPDMG, decreasing = TRUE),]
eco.dmg <- head(eco.ordered, 8)

#create plot
gp2 <- ggplot(eco.dmg, aes(x=EVTYPE, y=PROPDMG, fill=EVTYPE)) 
gp2 <- gp2 + geom_bar(stat="identity") 
gp2 <- gp2 + theme(axis.text.x = element_text(angle=45, hjust=1, size=8, color=2))
gp2 <- gp2 + geom_text(aes(x=eco.dmg$EVTYPE, y=eco.dmg$PROPDMG, angle=0, label=paste("$", floor(eco.dmg$PROPDMG), sep = ""), hjust=0.5, vjust=-0.5)) + theme(legend.position = "none")
gp2 <- gp2 + ylab("Property + Crop damage in dollars") + xlab("Event Type")
gp2

#show data used in a table
head(eco.dmg, 8)

Reproducible research - week 4 - Peer graded assignment 2 - weather data

synopsis

Data Processing

Results