Sysnopis

This report analyzes the NOAA Storm Database to answer:

1.Which event types (EVTYPE) are most harmful to population health?

  1. Which event types have the greatest economic consequences?

Data Processing

knitr::opts_chunk$set(echo = TRUE, message = FALSE, warning = FALSE)

Downloading data from the website:
[Storm Data](https://d396qusza40orc.cloudfront.net/repdata%2Fdata%2FStormData.csv.bz2)

or you can download file using the file URL

``` r
fileUrl <- "https://d396qusza40orc.cloudfront.net/repdata%2Fdata%2FStormData.csv.bz2"
fileZip <- "repdata-data-StormData.csv.bz2"

if (!file.exists(fileZip)) {
download.file(fileUrl, destfile = fileZip, mode = "wb")
}

data <- read.csv(fileZip)

Reading data from the raw file:

data <- read.csv(bzfile("repdata-data-StormData.csv.bz2"), header = TRUE, sep=",")

Question 1: Most harmful event type (population health)

Find the most harmful with respect to population health by FALITIES and INJURIES variable

harm <- data$FATALITIES + data$INJURIES
maxharm <- max(harm)
maxharm
## [1] 1742

Subsetting data to find the most harmful with respect to population health

mostharm <- data$EVTYPE[which(data$INJURIES + data$FATALITIES == maxharm)]
mostharm
## [1] "TORNADO"

Question 2: Which types of events have the greatest economic consequences?

Property Damage estimate in Billion dollar

prop <- data$PROPDMG[which(data$PROPDMGEXP == "B")]
prop
##  [1]   5.00   0.10   2.10   1.60   1.00   5.00   2.50   1.20   3.00   1.70
## [11]   3.00   1.50   5.15   1.00   1.04   2.50   5.42   1.30   4.83   4.00
## [21]   1.00   1.50  10.00  16.93  31.30   4.00   7.35  11.26   5.88   2.09
## [31] 115.00   1.00   4.00   1.50   1.80   1.00   1.50   2.80   1.00   2.00

The maximum property damage:

a <- max(prop)
a
## [1] 115

Crop Damage estimate in Billion dollar

crop <- data$CROPDMG[which(data$CROPDMGEXP == "B")]
crop
## [1] 0.40 5.00 0.50 0.20 5.00 1.51 1.00 0.00 0.00

The maximum property damage:

b <- max(crop)
b
## [1] 5

We can clearly see that the property damage is 115 billion dollar, which is the the largest amount.

The type of event which creates the greatest economic consequences:

econ <- data$EVTYPE[which((data$PROPDMG == a) & (data$PROPDMGEXP == "B"))]
econ
## [1] "FLOOD"

Plotting

Number of injuries and fatalities caused by Storm

tornado <- data[which(data$EVTYPE == "TORNADO"),]
library(ggplot2)

# create variable
data$total_human_cost <- data$FATALITIES + data$INJURIES

# aggregate to EVTYPE (so geom_bar(stat="identity") makes sense)
harm_plot <- aggregate(total_human_cost ~ EVTYPE, data = data, sum, na.rm = TRUE)
harm_plot <- harm_plot[order(-harm_plot$total_human_cost), ]
harm_plot10 <- harm_plot[1:10, ]

ggplot(harm_plot10, aes(x = reorder(EVTYPE, total_human_cost), y = total_human_cost)) +
  geom_col() +
  coord_flip() +
  xlab("Event type") +
  ylab("Total fatalities + injuries") +
  ggtitle("Top 10 Event Types by Total Human Cost")

The Property Damage caused by Flood

boxplot(data$PROPDMG[data$EVTYPE=="FLOOD"]
        , main ="Property Damage caused by Flood"
        ,xlab="FLOOD", ylab="Billions of Dollar")

Results

Question 1: Across the United States, which types of events (as indicated in the EVTYPE variable) are most harmful with respect to population health?

mostharm <- data$EVTYPE[which(data$INJURIES + data$FATALITIES == maxharm)]
mostharm
## [1] "TORNADO"

The answer is TORNADO.

Question 2: Across the United States, which types of events have the greatest economic consequences?

econ <- data$EVTYPE[which((data$PROPDMG == a) & (data$PROPDMGEXP == "B"))]
econ
## [1] "FLOOD"

The answer is FLOOD.