Analysis of adverse impacts on Health and Economy because of US Storms
Pentakota Nooshith
23/07/2020
1 : Synopsis
This project involves exploring the U.S. National Oceanic and Atmospheric Administration's (NOAA) storm database and explore various effects it had on both people and economy. The data in the database covers the from the year 1950 to the end of November 2011.
The below analysis provides information which type of severe events are harmful on:
1. Health
2. Economy
2 : Data Processing
2.1 : Data Loading
Download the compressed file and expand it.And then read the data into data.frame and convert it into data.table.
library("data.table")
library("ggplot2")
library("tidyr")
stormDT <- read.csv("repdata_data_StormData.csv")
#converting data.frame to data.table
stormDT <- as.data.table(stormDT)
#Dimensions of data
dim(stormDT)## [1] 902297 372.2 : Find the column in data.table
names(stormDT)## [1] "STATE__" "BGN_DATE" "BGN_TIME" "TIME_ZONE"
## [5] "COUNTY" "COUNTYNAME" "STATE" "EVTYPE"
## [9] "BGN_RANGE" "BGN_AZI" "BGN_LOCATI" "END_DATE"
## [13] "END_TIME" "COUNTY_END" "COUNTYENDN" "END_RANGE"
## [17] "END_AZI" "END_LOCATI" "LENGTH" "WIDTH"
## [21] "F" "MAG" "FATALITIES" "INJURIES"
## [25] "PROPDMG" "PROPDMGEXP" "CROPDMG" "CROPDMGEXP"
## [29] "WFO" "STATEOFFIC" "ZONENAMES" "LATITUDE"
## [33] "LONGITUDE" "LATITUDE_E" "LONGITUDE_" "REMARKS"
## [37] "REFNUM"2.3 : Data Subsetting
Here we take the required data columns which suit our specific requirement and subset them. From Documentation the following are the useful columns from dataset :
1. "EVTYPE"
2. "FATALITIES"
3. "INJURIES"
4. "PROPDMG"
5. "PROPDMGEXP"
6. "CROPDMG"
7. "CROPDMGEXP"
So here we subset data and create a new data table.
colsNeeded <- c("EVTYPE", "FATALITIES", "INJURIES", "PROPDMG",
"PROPDMGEXP", "CROPDMG", "CROPDMGEXP")
newStormDT <- stormDT[(INJURIES > 0 | FATALITIES > 0 | PROPDMG > 0 | CROPDMG > 0), c("EVTYPE", "FATALITIES", "INJURIES", "PROPDMG",
"PROPDMGEXP", "CROPDMG", "CROPDMGEXP")]2.4: Convert to Exponents
Since amount is shortformed to alphabet numbers let revert back to exponential or numveric forms. (K is thousands(10^3), M is millions(10^6), B is billions(10^9), and any other symbols are given 10^0 as the value )
# Lets Change all damage exponents to uppercase.
newStormDT[, c("PROPDMGEXP", "CROPDMGEXP")] <- lapply(newStormDT[, c("PROPDMGEXP", "CROPDMGEXP")], toupper)
# Find unique alphanumeric values
unique(newStormDT$PROPDMGEXP)## [1] "K" "M" "" "B" "+" "0" "5" "6" "4" "H" "2" "7" "3" "-"unique(newStormDT$CROPDMGEXP)## [1] "" "M" "K" "B" "?" "0"# Map damage alphanumeric exponents to numeric values.
dmgKey <- c("\"\"" = 10^0, "-" = 10^0, "+" = 10^0, "?" = 10^0, "0" = 10^0,
"1" = 10^1, "2" = 10^2, "3" = 10^3, "4" = 10^4,
"5" = 10^5, "6" = 10^6, "7" = 10^7, "8" = 10^8,
"9" = 10^9, "H" = 10^2, "K" = 10^3, "M" = 10^6, "B" = 10^9)
# Change the values in expenses to the exponential factors
newStormDT$PROPDMGEXP <- lapply(newStormDT$PROPDMGEXP, function(x) { dmgKey[as.character(x)] })
newStormDT[is.na(PROPDMGEXP), PROPDMGEXP := 10^0]
newStormDT$CROPDMGEXP <- lapply(newStormDT$CROPDMGEXP, function(x) { dmgKey[as.character(x)] })
newStormDT[is.na(CROPDMGEXP), CROPDMGEXP := 10^0]2.5 : Making cost columns
newStormDT$PROPCOST <- with(newStormDT, as.numeric(PROPDMGEXP) * PROPDMG)
newStormDT$CROPCOST <- with(newStormDT, as.numeric(CROPDMGEXP) * CROPDMG)2.6 : Calculating total cost loss based on EVTYPE
filteredCosts <- newStormDT[, .(PROPCOST = sum(PROPCOST), CROPCOST = sum(CROPCOST), TOTALCOST = sum(PROPCOST) + sum(CROPCOST)), by = .(EVTYPE)]
#Order the filtered data based on total costs
filteredCosts <- filteredCosts[order(-TOTALCOST), ]
#reduce the unncessary data
filteredCosts <- filteredCosts[with(filteredCosts, (PROPCOST > 0 & CROPCOST > 0))]
#lets put it down to top 10 data
filteredCosts <- filteredCosts[1:10,]
# Look at some top data
head(filteredCosts, 10)## EVTYPE PROPCOST CROPCOST TOTALCOST
## 1: FLOOD 144657709807 5661968450 150319678257
## 2: HURRICANE/TYPHOON 69305840000 2607872800 71913712800
## 3: TORNADO 56947380677 414953270 57362333947
## 4: STORM SURGE 43323536000 5000 43323541000
## 5: HAIL 15735267513 3025954473 18761221986
## 6: FLASH FLOOD 16822673979 1421317100 18243991079
## 7: DROUGHT 1046106000 13972566000 15018672000
## 8: HURRICANE 11868319010 2741910000 14610229010
## 9: RIVER FLOOD 5118945500 5029459000 10148404500
## 10: ICE STORM 3944927860 5022113500 89670413602.7 : Calculating total Fatalities and Injuries
filteredFI <- newStormDT[, .(FATALITIES = sum(FATALITIES), INJURIES = sum(INJURIES), TOTAL = sum(FATALITIES) + sum(INJURIES)), by = .(EVTYPE)]
#Order the filtered data based on fatalities
filteredFI <- filteredFI[order(-FATALITIES), ]
#reduce the unncessary data
filteredFI <- filteredFI[with(filteredFI, (FATALITIES > 0 & INJURIES > 0))]
#lets put it down to top 10 data
filteredFI <- filteredFI[1:10,]
# Look at some top data
head(filteredFI, 10)## EVTYPE FATALITIES INJURIES TOTAL
## 1: TORNADO 5633 91346 96979
## 2: EXCESSIVE HEAT 1903 6525 8428
## 3: FLASH FLOOD 978 1777 2755
## 4: HEAT 937 2100 3037
## 5: LIGHTNING 816 5230 6046
## 6: TSTM WIND 504 6957 7461
## 7: FLOOD 470 6789 7259
## 8: RIP CURRENT 368 232 600
## 9: HIGH WIND 248 1137 1385
## 10: AVALANCHE 224 170 3943: Results
3.1 : Events that are Most Harmful to Population Health
# Gather the data to fit into graph model
filteredFI <- gather(filteredFI, HEALTH, VALUE, FATALITIES:TOTAL)
healthChart <- ggplot(filteredFI, aes(x = reorder(EVTYPE, -VALUE), y = VALUE))
healthChart <- healthChart + geom_bar(stat = "identity", aes(fill = HEALTH), position = "dodge")
healthChart <- healthChart + labs(x = "Type of Event", y = "Count", title = "TOP 10 Harmful Events in US")
healthChart <- healthChart + theme(axis.text.x = element_text(angle = 50, hjust = 1), plot.title = element_text(hjust = 0.5))
healthChart
plot of chunk gather1
3.2 : Events that are Most Harmful to Economy
# Gather the data to fit into graph model
filteredCosts <- gather(filteredCosts, DMGTYPE, VALUE, PROPCOST:TOTALCOST)
economyChart <- ggplot(filteredCosts, aes(x = reorder(EVTYPE, -VALUE), y = VALUE))
economyChart <- economyChart + geom_bar(stat = "identity", aes(fill = DMGTYPE), position = "dodge")
economyChart <- economyChart + labs(x = "Type of Event", y = "Cost(in Dollars)", title = "TOP 10 Economic Consequences in US Storm Events")
economyChart <- economyChart + theme(axis.text.x = element_text(angle = 50, hjust = 1), plot.title = element_text(hjust = 0.5))
economyChart
plot of chunk gather2