Coursera - Reproducible Research - Project 2

1 Synopsis

This analysis is for the Coursera Reproducible Research course Week 4 Project 2, part of the Data Science Specialization. The project is to explore the impact of severe weather events on human heatlth (injuries and fatalities) and finanical damage (property and crop damage)

2 Data Processing

library(tidyverse)
library(grid)
library(gridExtra)
library(pander)

NOAA <- read_csv("repdata-data-StormData.csv")

names(NOAA)

##  [1] "STATE__"    "BGN_DATE"   "BGN_TIME"   "TIME_ZONE"  "COUNTY"    
##  [6] "COUNTYNAME" "STATE"      "EVTYPE"     "BGN_RANGE"  "BGN_AZI"   
## [11] "BGN_LOCATI" "END_DATE"   "END_TIME"   "COUNTY_END" "COUNTYENDN"
## [16] "END_RANGE"  "END_AZI"    "END_LOCATI" "LENGTH"     "WIDTH"     
## [21] "F"          "MAG"        "FATALITIES" "INJURIES"   "PROPDMG"   
## [26] "PROPDMGEXP" "CROPDMG"    "CROPDMGEXP" "WFO"        "STATEOFFIC"
## [31] "ZONENAMES"  "LATITUDE"   "LONGITUDE"  "LATITUDE_E" "LONGITUDE_"
## [36] "REMARKS"    "REFNUM"

stormData <- NOAA[,c("EVTYPE", "FATALITIES", "INJURIES", "PROPDMG",
                     "PROPDMGEXP","CROPDMG","CROPDMGEXP")]

2.1 Data Processing for Health Analysis

Health <- stormData %>% group_by(EVTYPE) %>% summarise(Injury = sum(INJURIES), Fatality =sum(FATALITIES))

HealthInjury <- Health [order (Health$Injury, decreasing = T), -3]

HealthFatality <- Health [order (Health$Fatality, decreasing = T), -2]

2.2 Data Processing for Economic Damage Analysis

# check PROPDMGEXP class

class(stormData$PROPDMGEXP)

## [1] "character"

# convert to factor

stormData$PROPDMGEXP <- as.factor(stormData$PROPDMGEXP)

stormData$CROPDMGEXP <- as.factor(stormData$CROPDMGEXP)

# subset stormData

ED <- stormData[,-c(2,3)]

# check PROPDMGEXP levels

levels(ED$PROPDMGEXP)

##  [1] "-" "?" "+" "0" "1" "2" "3" "4" "5" "6" "7" "8" "B" "h" "H" "K" "m"
## [18] "M"

table(factor(ED$PROPDMGEXP))

## 
##      -      ?      +      0      1      2      3      4      5      6 
##      1      8      5    216     25     13      4      4     28      4 
##      7      8      B      h      H      K      m      M 
##      5      1     40      1      6 424665      7  11330

sum(is.na(ED$PROPDMGEXP))

## [1] 465934

ED$Prop.DMG[ED$PROPDMGEXP %in% c("-","?","+")] <- 0

## Warning: Unknown or uninitialised column: 'Prop.DMG'.

ED$Prop.DMG[ED$PROPDMGEXP %in% c("h","H")] <- 100

ED$Prop.DMG[ED$PROPDMGEXP == "0"] <- 1

ED$Prop.DMG[ED$PROPDMGEXP == ""] <- 0

ED$Prop.DMG[ED$PROPDMGEXP == "1"] <- 10

ED$Prop.DMG[ED$PROPDMGEXP == "2"] <- 100

ED$Prop.DMG[ED$PROPDMGEXP == "3"] <- 1000

ED$Prop.DMG[ED$PROPDMGEXP == "4"] <- 1*10^4

ED$Prop.DMG[ED$PROPDMGEXP == "5"] <- 1*10^5

ED$Prop.DMG[ED$PROPDMGEXP == "6"] <- 1*10^6

ED$Prop.DMG[ED$PROPDMGEXP == "7"] <- 1* 10^7

ED$Prop.DMG[ED$PROPDMGEXP == "8"] <- 1* 10^8

ED$Prop.DMG[ED$PROPDMGEXP == "K"] <- 1000

ED$Prop.DMG[ED$PROPDMGEXP == "B"] <- 1*10^9

ED$Prop.DMG[ED$PROPDMGEXP %in% c("m","M")] <- 1*10^6

sum(is.na(ED$Prop.DMG))

## [1] 465934

Property <- ED[, c(1,2,6)]

levels(ED$CROPDMGEXP)

## [1] "?" "0" "2" "B" "k" "K" "m" "M"

table(factor(ED$CROPDMGEXP))

## 
##      ?      0      2      B      k      K      m      M 
##      7     19      1      9     21 281832      1   1994

ED$Crop.DMG[ED$CROPDMGEXP == "?"] <- 0

## Warning: Unknown or uninitialised column: 'Crop.DMG'.

ED$Crop.DMG[ED$CROPDMGEXP == ""] <- 0

ED$Crop.DMG[ED$CROPDMGEXP == "0"] <- 0

ED$Crop.DMG[ED$CROPDMGEXP == "2"] <- 100

ED$Crop.DMG[ED$CROPDMGEXP == "B"] <- 1*10^9

ED$Crop.DMG[ED$CROPDMGEXP %in% c("k","K")] <- 1000

ED$Crop.DMG[ED$CROPDMGEXP %in% c("m","M")] <- 1*10^6

Crop <- ED[,c(1,4,7)]

3 Results

3.1 Health Impact

Top 6 weather events cause injuries.

pander(head(HealthInjury,6))

EVTYPE	Injury
TORNADO	91346
TSTM WIND	6957
FLOOD	6789
EXCESSIVE HEAT	6525
LIGHTNING	5230
HEAT	2100

Top 6 weather events cause fatalities.

pander(head(HealthFatality,6))

EVTYPE	Fatality
TORNADO	5633
EXCESSIVE HEAT	1903
FLASH FLOOD	978
HEAT	937
LIGHTNING	816
TSTM WIND	504

3.2 Economic Damage

Property$DMG.Value <- Property$PROPDMG*Property$Prop.DMG

Crop$DMG.Value <- Crop$CROPDMG*Crop$Crop.DMG

PropertyDMD <- Property %>% group_by(EVTYPE) %>% 
    summarise(Damage = sum(DMG.Value)) %>% 
    arrange(desc(Damage))

PropertyDMDTop5 <- PropertyDMD [c(1:5),]

PropertyDMDTop5$Damage <- PropertyDMDTop5$Damage/10^9

names(PropertyDMDTop5) <- c("Events","Billions")

CropDMD <- Crop %>% group_by(EVTYPE) %>% 
    summarise(Damage = sum(DMG.Value)) %>% 
    arrange(desc(Damage))

CropDMDTop5 <- CropDMD[c(1:5),]

CropDMDTop5$Damage <- CropDMDTop5$Damage/10^9

names(CropDMDTop5) <- c("Events","Billions")

pander(PropertyDMDTop5)

Events	Billions
TORNADOES, TSTM WIND, HAIL	1.6
WILD FIRES	0.6241
HAILSTORM	0.241
HIGH WINDS/COLD	0.1105
River Flooding	0.1062

pander(CropDMDTop5)

Events	Billions
EXCESSIVE WETNESS	0.142
COLD AND WET CONDITIONS	0.066
Early Frost	0.042
Damaging Freeze	0.03413
Freeze	0.0105

Top 5 weather events on Economic Impact

g1 <- ggplot(PropertyDMDTop5,aes(x=Events, y = Billions)) + geom_col()+
    labs(title = "Property Damage (Top 5)")

g2 <- ggplot(CropDMDTop5,aes(x=Events, y = Billions)) + geom_col()+
    labs(title = " Crop Damage (Top 5)")

grid.arrange(g1,g2,nrow = 2)