Weather Event Data Analysis
Synopsis
Extream weather such as storm can damage on the crops and properties, and cause substantial human injury and fatality. Properly tracking and analysis of the historical weather events data might provide insight on how to avoid or reduce its detrimental effects. This report attempts to analyze the weather event data collected from 1950 to 2011 accross the United States. Specifically, this report focus on the most harmful weather events on the public health as well as the economy. To evaluate the public health effect, fatalities and injuries were choosed. This report found that Torando is the most harmful weather event types for both fatalities and injuries. To evaluate the economic effect, propertaties damage and crops damage were selected. This report found that Torando is also the most harmful event types for propertaties. Among others, Hail was the most dangerious event types for crops.
Data Processing
load package:
library(knitr)
library(dplyr)##
## Attaching package: 'dplyr'
##
## The following object is masked from 'package:stats':
##
## filter
##
## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionlibrary(lattice)Read partially of the weather event data which is in a CSV format with proper header
stormdata<-read.table("repdata-data-StormData.csv", head=TRUE, sep=",",nrows=100)Briefly checking the data how does it look like
head(stormdata, n=2L)## STATE__ BGN_DATE BGN_TIME TIME_ZONE COUNTY COUNTYNAME STATE
## 1 1 4/18/1950 0:00:00 130 CST 97 MOBILE AL
## 2 1 4/18/1950 0:00:00 145 CST 3 BALDWIN AL
## EVTYPE BGN_RANGE BGN_AZI BGN_LOCATI END_DATE END_TIME COUNTY_END
## 1 TORNADO 0 NA NA NA NA 0
## 2 TORNADO 0 NA NA NA NA 0
## COUNTYENDN END_RANGE END_AZI END_LOCATI LENGTH WIDTH F MAG FATALITIES
## 1 NA 0 NA NA 14 100 3 0 0
## 2 NA 0 NA NA 2 150 2 0 0
## INJURIES PROPDMG PROPDMGEXP CROPDMG CROPDMGEXP WFO STATEOFFIC ZONENAMES
## 1 15 25.0 K 0 NA NA NA NA
## 2 0 2.5 K 0 NA NA NA NA
## LATITUDE LONGITUDE LATITUDE_E LONGITUDE_ REMARKS REFNUM
## 1 3040 8812 3051 8806 NA 1
## 2 3042 8755 0 0 NA 2Read all data interested and preprocessing the data. All missing values were labeled as “NA”
alldata<-subset(read.table("repdata-data-StormData.csv", head=TRUE, na.strings="NA",sep=","), select=c("BGN_DATE","EVTYPE", "FATALITIES", "INJURIES", "PROPDMG", "CROPDMG"))
#Create a new variable "YEAR"
tempdate<-as.Date(strptime(alldata$BGN_DATE, '%m/%d/%Y'))
tempyear<-format(tempdate, '%Y')
alldata$YEAR<-as.factor(tempyear)
alldata<-alldata[, c("EVTYPE", "FATALITIES", "INJURIES", "PROPDMG", "CROPDMG", "YEAR")]
dim(alldata)## [1] 902297 6summary(alldata)## EVTYPE FATALITIES INJURIES
## HAIL :288661 Min. : 0.0000 Min. : 0.0000
## TSTM WIND :219940 1st Qu.: 0.0000 1st Qu.: 0.0000
## THUNDERSTORM WIND: 82563 Median : 0.0000 Median : 0.0000
## TORNADO : 60652 Mean : 0.0168 Mean : 0.1557
## FLASH FLOOD : 54277 3rd Qu.: 0.0000 3rd Qu.: 0.0000
## FLOOD : 25326 Max. :583.0000 Max. :1700.0000
## (Other) :170878
## PROPDMG CROPDMG YEAR
## Min. : 0.00 Min. : 0.000 2011 : 62174
## 1st Qu.: 0.00 1st Qu.: 0.000 2008 : 55663
## Median : 0.00 Median : 0.000 2010 : 48161
## Mean : 12.06 Mean : 1.527 2009 : 45817
## 3rd Qu.: 0.50 3rd Qu.: 0.000 2006 : 44034
## Max. :5000.00 Max. :990.000 2007 : 43289
## (Other):603159#Check for missing value
colSums(is.na(alldata))## EVTYPE FATALITIES INJURIES PROPDMG CROPDMG YEAR
## 0 0 0 0 0 0Results
The most harmful weather event for the public health
Events cause fatalities or injuries
In this report, two variables, fatalities and injuries, were choosed for evaluating the public health effects caused by weather. To collect meaningful data, the report will subset those events which did not cause negative loss with respect to the fatalities as well as injuries.
EventHealth<-subset(alldata, FATALITIES>=0|INJURIES>=0, select=c("YEAR","EVTYPE","FATALITIES", "INJURIES"))
dim(EventHealth)## [1] 902297 4Event which caused maximal one-time injuries or fatalities
##The event caused maximal one-time injuries
subset(EventHealth, INJURIES==max(EventHealth$INJURIES) )## YEAR EVTYPE FATALITIES INJURIES
## 157885 1979 TORNADO 42 1700##The event caused maximal one-time fatalities
subset(EventHealth, FATALITIES==max(EventHealth$FATALITIES) )## YEAR EVTYPE FATALITIES INJURIES
## 198704 1995 HEAT 583 0Event type which caused most injuries or fatalities in total
#Event type which caused most injuries in total
gyinj<-group_by(EventHealth, EVTYPE)
gyinjutotal<-summarize(gyinj, INJURIES=sum(INJURIES))
gyinjutotalmax<-subset(gyinjutotal, INJURIES==max(gyinjutotal$INJURIES))
gyinjutotalmax## Source: local data frame [1 x 2]
##
## EVTYPE INJURIES
## 1 TORNADO 91346#Event type which caused most fatalities in total
gyfatatotal<-summarize(gyinj, avgfata=sum(FATALITIES))
gyfatatotalmax<-subset(gyfatatotal, avgfata==max(gyfatatotal$avgfata))
gyfatatotalmax## Source: local data frame [1 x 2]
##
## EVTYPE avgfata
## 1 TORNADO 5633The injuries caused by Tornado and other event type from 1950 to 2011
The summary of injuries caused by Tornado and Other Weather Types
#Create a new variable which consists of two levels
wt<-factor((EventHealth$EVTYPE==gyinjutotalmax$EVTYPE),
levels = c(FALSE, TRUE), labels= c("Other","Tornado"))
EventHealth$WEATHERTYPE<-wt
#Average Injuries
grp<-group_by(EventHealth, WEATHERTYPE)
im<-summarize(grp, INJURIES=mean(INJURIES))
barchart(INJURIES~WEATHERTYPE, data=im, col= "blue"
,main="Average Injuries", xlab="Weather Type"
,ylab="Average Injuries")
#Total Fatalities
it<-summarize(grp, INJURIES=sum(INJURIES))
barchart(INJURIES~WEATHERTYPE, data=it, col= "blue"
,main="Total Injuries", xlab="Weather Type"
,ylab="Total Injuries")
The detailed injuriess caused by Tornado and Other Weather Types
injusub<-EventHealth[,c("YEAR","WEATHERTYPE","INJURIES")]
grp_cols<-names(injusub)[-3]
dots<-lapply(grp_cols, as.symbol)
sumdata<-injusub %>% group_by_(.dots=dots) %>% summarize(INJURIES=sum(INJURIES))
barchart(INJURIES~WEATHERTYPE|YEAR, data=sumdata, strip=TRUE, pch=20,main="Injuries caused by Tornado and Other Event Types", xlab="Weather Type", ylab="Total Injuries")
The fatalities caused by Tornado and other event type from 1950 to 2011
Summary of fatalities caused by Tornado and Other Weather Types
wt<-factor((EventHealth$EVTYPE==gyfatatotalmax$EVTYPE),
levels = c(FALSE, TRUE), labels= c("Other","Tornado"))
EventHealth$WEATHERTYPE<-wt
#Average Fatalities
grp<-group_by(EventHealth, WEATHERTYPE)
fm<-summarize(grp, FATALITIES=mean(FATALITIES))
barchart(FATALITIES~WEATHERTYPE, data=fm, col= "blue"
,main="Average Fatalities", xlab="Weather Type"
,ylab="Average Fatalities")
#Total Fatalities
ft<-summarize(grp, FATALITIES=sum(FATALITIES))
barchart(FATALITIES~WEATHERTYPE, data=ft, col= "blue"
,main="Total Fatalities", xlab="Weather Type"
,ylab="Total Fatalities")
The detailed fatalities caused by Tornado and Other Weather Types
injusub<-EventHealth[,c("YEAR","WEATHERTYPE","FATALITIES")]
grp_cols<-names(injusub)[-3]
dots<-lapply(grp_cols, as.symbol)
sumdata<-injusub %>% group_by_(.dots=dots) %>% summarize(FATALITIES=sum(FATALITIES))
barchart(FATALITIES~WEATHERTYPE|YEAR, data=sumdata, strip=TRUE, pch=20,main="Fatalities caused by Tornado and Other Event Types", xlab="Weather Type", ylab="Total Fatalities")
The most harmful weather event for the economy
Events cause properties damage or crops damage
In this report, two variables, properties damage and crops damage, were choosed for evaluating the economic effects caused by weather. To collect meaningful data, the report will subset those events which did not cause negative loss with respect to the properties damage or crops damage.
EventEconomy<-subset(alldata, PROPDMG>=0|CROPDMG>=0, select=c("YEAR","EVTYPE","PROPDMG", "CROPDMG"))
dim(EventEconomy)## [1] 902297 4Event type which caused maximal one-time property or crop damages
##The event caused maximal one-time property damages
subset(EventEconomy, PROPDMG==max(EventEconomy$PROPDMG) )## YEAR EVTYPE PROPDMG CROPDMG
## 778568 2009 THUNDERSTORM WIND 5000 0
## 808182 2010 FLASH FLOOD 5000 0
## 808183 2010 FLASH FLOOD 5000 0
## 900685 2011 WATERSPOUT 5000 0##The event caused maximal one-time crop damages
subset(EventEconomy, CROPDMG==max(EventEconomy$CROPDMG) )## YEAR EVTYPE PROPDMG CROPDMG
## 544253 2004 DROUGHT 0 990Event type which caused most properties damage or crops damage in total
#Event type which caused properties damage in total
gyinj<-group_by(EventEconomy, EVTYPE)
gyproptotal<-summarize(gyinj, PROPDMG=sum(PROPDMG))
gyproptotalmax<-subset(gyproptotal, PROPDMG==max(gyproptotal$PROPDMG))
gyproptotalmax## Source: local data frame [1 x 2]
##
## EVTYPE PROPDMG
## 1 TORNADO 3212258#Event type which caused crops damage in total
gycroptotal<-summarize(gyinj, CROPDMG=sum(CROPDMG))
gycroptotalmax<-subset(gycroptotal, CROPDMG==max(gycroptotal$CROPDMG))
gycroptotalmax## Source: local data frame [1 x 2]
##
## EVTYPE CROPDMG
## 1 HAIL 579596.3The property damage caused by Tornado and other event type from 1950 to 2011
#Create a new variable which consists of two levels
wt<-factor((EventEconomy$EVTYPE==gyproptotalmax$EVTYPE),
levels = c(FALSE, TRUE), labels= c("Other","Tornado"))
EventEconomy$WEATHERTYPE<-wt
#Average Property Damage
grp<-group_by(EventEconomy, WEATHERTYPE)
pm<-summarize(grp, PROPDMG=mean(PROPDMG))
barchart(PROPDMG~WEATHERTYPE, data=pm, col= "blue"
,main="Average Property Damage", xlab="Weather Type"
,ylab="Average Property Damage")
#Total Property Damage
pt<-summarize(grp, PROPDMG=sum(PROPDMG))
barchart(PROPDMG~WEATHERTYPE, data=pt, col= "blue"
,main="Total Property Damage", xlab="Weather Type"
,ylab="Total Property Damage")
The detail property damage caused by Tornado and Other Event Types
injusub<-EventEconomy[,c("YEAR","WEATHERTYPE","PROPDMG")]
grp_cols<-names(injusub)[-3]
dots<-lapply(grp_cols, as.symbol)
sumdata<-injusub %>% group_by_(.dots=dots) %>% summarize(PROPDMG=sum(PROPDMG))
barchart(PROPDMG~WEATHERTYPE|YEAR, data=sumdata, strip=TRUE, pch=20,main="Property Damage Caused by Tornado and Other Event Types", xlab="Weather Type", ylab="Total Injuries")
The crop damage caused by Hail and other event type from 1950 to 2011
The average crop damage caused by Hail and Other Event Types
wt<-factor((EventEconomy$EVTYPE==gycroptotalmax$EVTYPE),
levels = c(FALSE, TRUE), labels= c("Other","Hail"))
EventEconomy$WEATHERTYPE<-wt
#Average Crop Damage
grp<-group_by(EventEconomy, WEATHERTYPE)
cm<-summarize(grp, CROPDMG=mean(CROPDMG))
barchart(CROPDMG~WEATHERTYPE, data=cm, col= "blue"
,main="Average Crop Damage", xlab="Weather Type"
,ylab="Average Crop Damage")
#Total Crop Damage
ct<-summarize(grp, CROPDMG=sum(CROPDMG))
barchart(CROPDMG~WEATHERTYPE, data=ct, col= "blue"
,main="Total Crop Damage", xlab="Weather Type"
,ylab="Total Crop Damage")
The detail crop damage caused by Hail and Other Event Types
injusub<-EventEconomy[,c("YEAR","WEATHERTYPE","CROPDMG")]
grp_cols<-names(injusub)[-3]
dots<-lapply(grp_cols, as.symbol)
sumdata<-injusub %>% group_by_(.dots=dots) %>% summarize(CROPDMG=sum(CROPDMG))
barchart(CROPDMG~WEATHERTYPE|YEAR, data=sumdata, strip=TRUE, pch=20,main="Crop Damage Caused by Hail and Other Event Types", xlab="Weather Type", ylab="Total Crop Damage")