Synopsis: Utilizing R Markdown to document the process of addressing project questions. Due to the large size of the data frame repdata_data_StormData.csv, the initial step involves splitting the data and extracting only the required information. The code first loads the data, then creates a new data frame containing the relevant columns “FATALITIES”, “INJURIES”, and “EVTYPE”. The EVTYPE column contains numerous instances of weather event types with varying recorded injuries and fatalities. Using this information, I identify the unique weather types and calculate the total fatalities and injuries for each type, storing the results in a new data frame for further analysis. The final step involves extracting the necessary information to address the project questions.
Loading the data: Loading the data using read.csv with the document file_path as the input
#Load the data
file_path <- "Downloads/repdata_data_StormData.csv"
data <- read.csv(file_path)#Create a new data frame with only relevant columns (FATALITIES, INJURIES, EVTYPE)
new_df <- data[, c("FATALITIES", "INJURIES", "EVTYPE")]
head(new_df)## FATALITIES INJURIES EVTYPE
## 1 0 15 TORNADO
## 2 0 0 TORNADO
## 3 0 2 TORNADO
## 4 0 2 TORNADO
## 5 0 2 TORNADO
## 6 0 6 TORNADO#Sum up all the recorded injuries and fatalities with the unique event type
unique_events <- unique(new_df$EVTYPE)
# Initialize an empty data frame to store the results
event_totals <- data.frame(EVTYPE = character(), FATALITIES = numeric(), INJURIES = numeric(), stringsAsFactors = FALSE)
# Loop through each unique event type
for (event in unique_events) {
# Subset the data for the current event type
subset_data <- new_df[new_df$EVTYPE == event, ]
# Calculate total fatalities and injuries
total_fatalities <- sum(subset_data$FATALITIES, na.rm = TRUE)
total_injuries <- sum(subset_data$INJURIES, na.rm = TRUE)
# Add the totals to the results data frame
event_totals <- rbind(event_totals, data.frame(EVTYPE = event, FATALITIES = total_fatalities, INJURIES = total_injuries))
}
# Print the result
head(event_totals)## EVTYPE FATALITIES INJURIES
## 1 TORNADO 5633 91346
## 2 TSTM WIND 504 6957
## 3 HAIL 15 1361
## 4 FREEZING RAIN 7 23
## 5 SNOW 5 29
## 6 ICE STORM/FLASH FLOOD 0 2event_fatalities_order <- event_totals[order(-event_totals$FATALITIES), ]
event_injuries_order <- event_totals[order(-event_totals$INJURIES), ]
head(event_fatalities_order)## EVTYPE FATALITIES INJURIES
## 1 TORNADO 5633 91346
## 99 EXCESSIVE HEAT 1903 6525
## 20 FLASH FLOOD 978 1777
## 27 HEAT 937 2100
## 15 LIGHTNING 816 5230
## 2 TSTM WIND 504 6957head(event_injuries_order)## EVTYPE FATALITIES INJURIES
## 1 TORNADO 5633 91346
## 2 TSTM WIND 504 6957
## 36 FLOOD 470 6789
## 99 EXCESSIVE HEAT 1903 6525
## 15 LIGHTNING 816 5230
## 27 HEAT 937 2100Results Question 1: event_fatalities_order <- event_totals[order(-event_totals$FATALITIES), ] Using the command to order the weather event with the most fatalities, results can be interpreted from the table.
event_injuries_order <- event_totals[order(-event_totals$INJURIES), ] Using the command to order the weather event with the most injuries, results can be interpreted from the table.
# Create a new data frame with only relevant columns (PROPDMG, PROPDMGEXP, CROPDMG, EVTYPE)
new_df1 <- data[, c("PROPDMG", "PROPDMGEXP", "CROPDMG", "EVTYPE")]
# Function to convert PROPDMGEXP to numeric multiplier, using PROPDMGEXP for CROPDMG
exp_to_multiplier <- function(exp) {
multiplier <- switch(exp,
"K" = 1e3,
"M" = 1e6,
1)
return(multiplier)
}
# Convert the letters in PROPDMGEXP to numeric multipliers
multipliers <- sapply(new_df1$PROPDMGEXP, exp_to_multiplier)
# Multiply PROPDMG, by the multipliers to get total property damage
new_df1$total_prop_damage <- new_df1$PROPDMG * multipliers
new_df1$total_crop_damage <- new_df1$CROPDMG * multipliers
# Determine unique event types in EVTYPE
unique_events <- unique(new_df1$EVTYPE)
# Initialize an empty data frame to store the results
event_totals1 <- data.frame(EVTYPE = character(),
total_prop_damage = numeric(),
total_crop_damage = numeric(),
stringsAsFactors = FALSE)
# Loop through each unique event type
for (event in unique_events) {
# Subset the data for the current event type
subset_data <- new_df1[new_df1$EVTYPE == event, ]
# Calculate total property damage and crop damage
total_prop_damage <- sum(subset_data$total_prop_damage, na.rm = TRUE)
total_crop_damage <- sum(subset_data$total_crop_damage, na.rm = TRUE)
# Add the totals to the results data frame
event_totals1 <- rbind(event_totals1, data.frame(EVTYPE = event,
total_prop_damage = total_prop_damage,
total_crop_damage = total_crop_damage))
}
# Add a new column for total damage
event_totals1$total_damage <- event_totals1$total_prop_damage + event_totals1$total_crop_damage
head(event_totals1)## EVTYPE total_prop_damage total_crop_damage total_damage
## 1 TORNADO 51625660796 28269878512 79895539307
## 2 TSTM WIND 4484928495 7684657862 12169586357
## 3 HAIL 13927367054 15314323923 29241690977
## 4 FREEZING RAIN 8111500 0 8111500
## 5 SNOW 14762550 10000 14772550
## 6 ICE STORM/FLASH FLOOD 0 0 0# Order Total_Damage from highest to lowest
event_total_damage <- event_totals1[order(-event_totals1$total_damage), ]
head(event_total_damage)## EVTYPE total_prop_damage total_crop_damage total_damage
## 1 TORNADO 51625660796 28269878512 79895539307
## 36 FLOOD 22157709930 53751978402 75909688331
## 20 FLASH FLOOD 15140812068 38822140319 53962952387
## 3 HAIL 13927367054 15314323923 29241690977
## 2 TSTM WIND 4484928495 7684657862 12169586357
## 46 HIGH WIND 3970046296 7174066263 11144112559Results Question 2: event_total_damage <- event_totals1[order(-event_totals1$total_damage), ] Using the command to order the weather event with the most fatalities, results can be interpreted from the table.
Conclusion: After analyzing the data, it is evident that tornadoes caused the highest number of fatalities and injuries, as well as the most damage to property and crops.