Synopsis

Storms and other severe weather events can cause both public health and economic problems for communities and municipalities. Many severe events can result in fatalities, injuries, and property damage, and preventing such outcomes to the extent possible is a key concern.

This project involves exploring the U.S. National Oceanic and Atmospheric Administration’s (NOAA) storm database. This database tracks characteristics of major storms and weather events in the United States, including when and where they occur, as well as estimates of any fatalities, injuries, and property damage.

The data analysis addresses the following questions :
1. Across the United States, which types of events (as indicated in the EVTYPEEVTYPE variable) are most harmful with respect to population health?
2. Across the United States, which types of events have the greatest economic consequences?

Tornados are the most harmfull events on population health, whereas floods are responsible for the most economic damage.

Data Processing

Data Processing describes (in words and code) how the data were loaded into R and processed for analysis. In particular, my analysis must start from the raw CSV file containing the data.

Loading packages

library(ggplot2)
library(reshape2)
library(magrittr)
library(dplyr)
## 
## Attaching package: 'dplyr'
## The following objects are masked from 'package:stats':
## 
##     filter, lag
## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union

Loading the data

data_NOAA <- read.csv("./repdata_data_StormData.csv", sep = ",", header = T)

Subsetting data

data_NOAA <- data_NOAA[, c('EVTYPE','FATALITIES','INJURIES', 'PROPDMG', 'PROPDMGEXP', 'CROPDMG', 'CROPDMGEXP')]

Cleaning data

data_cleaned <- data_NOAA %>% 
                mutate(CROPDMG = CROPDMG * case_when(
                        CROPDMGEXP == "B" ~ 10^9,
                        CROPDMGEXP == "k" | CROPDMGEXP == "K" ~ 10^3,
                        CROPDMGEXP == "m" | CROPDMGEXP == "M" ~ 10^6,
                        CROPDMGEXP == 2 ~ 2,
                        TRUE ~ 0), CROPDMGEXP = NULL)

data_cleaned %<>% mutate(PROPDMG = PROPDMG * case_when(
                        PROPDMGEXP == "B" ~ 10^9,
                        PROPDMGEXP == "k" | PROPDMGEXP == "K" ~ 10^3,
                        PROPDMGEXP == "m" | PROPDMGEXP == "M" ~ 10^6,
                        PROPDMGEXP == "h" | PROPDMGEXP == "H" ~ 10^2,
                        PROPDMGEXP == 1 ~ 1,
                        PROPDMGEXP == 2 ~ 2,
                        PROPDMGEXP == 3 ~ 3,
                        PROPDMGEXP == 4 ~ 4,
                        PROPDMGEXP == 5 ~ 5,
                        PROPDMGEXP == 6 ~ 6,
                        PROPDMGEXP == 7 ~ 7,
                        PROPDMGEXP == 8 ~ 8,
                        TRUE ~ 0), PROPDMGEXP = NULL)

data_final <- melt(data_cleaned, id = c('EVTYPE'), variable.name = "category", value.name = "value", na.rm = TRUE)

Structure

str(data_final)
## 'data.frame':    3609188 obs. of  3 variables:
##  $ EVTYPE  : Factor w/ 985 levels "   HIGH SURF ADVISORY",..: 834 834 834 834 834 834 834 834 834 834 ...
##  $ category: Factor w/ 4 levels "FATALITIES","INJURIES",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ value   : num  0 0 0 0 0 0 0 0 1 0 ...

Results

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

fatalities <- data_final[data_final$category == "FATALITIES", ]
fat_nr <- aggregate(value ~ EVTYPE, data = fatalities, sum)
fat_nr <- fat_nr[order(-fat_nr$value), ][1:15, ]
fat_nr$EVTYPE <- factor(fat_nr$EVTYPE, levels = fat_nr$EVTYPE)

# Plot nr. of fatalities with the most harmful event type
ggplot(fat_nr, aes(x = EVTYPE, y = value)) + 
    geom_bar(stat = "identity", fill = "blue") + 
    theme(axis.text.x = element_text(angle = 90, hjust = 1)) + 
    xlab("Event Type") + ylab("Fatalities") + ggtitle("Number of fatalities for the most 15 harmful weather events")

injuries <- data_final[data_final$category == "INJURIES", ]
inj_nr <- aggregate(value ~ EVTYPE, data = injuries, sum)
inj_nr <- inj_nr[order(-inj_nr$value), ][1:15, ]
inj_nr$EVTYPE <- factor(inj_nr$EVTYPE, levels = inj_nr$EVTYPE)

# Plot nr. of injuries with the most harmful event type
ggplot(inj_nr, aes(x = EVTYPE, y = value)) + 
    geom_bar(stat = "identity", fill = "blue") + 
    theme(axis.text.x = element_text(angle = 90, hjust = 1)) + 
    xlab("Event Type") + ylab("Injuries") + ggtitle("Number of injuries for the  most 15 harmful weather events")

  1. Across the United States, which types of events have the greatest economic consequences ?
damages <- data_final[data_final$category == "PROPDMG|CROPDMG", ]
dam_nr <- aggregate(value ~ EVTYPE, data = data_final, sum)
names(dam_nr) = c("EVTYPE", "DAMAGE")
dam_nr <- dam_nr[order(-dam_nr$DAMAGE), ][1:15, ]
dam_nr$EVTYPE <- factor(dam_nr$EVTYPE, levels = dam_nr$EVTYPE)

# Plot nr. of damages with the most harmful event type
ggplot(dam_nr, aes(x = EVTYPE, y = DAMAGE)) + 
    geom_bar(stat = "identity", fill = "blue") + 
    theme(axis.text.x = element_text(angle = 90, hjust = 1)) + 
    xlab("Event Type") + ylab("Damages") + ggtitle("Property & Crop damages for the most 15 weather events")

As seen in the picture, floods cause the most damage.