## Analysis of Federal Disaster Closeout Durations
Introduction
The Federal Emergency Management Agency (FEMA) maintains the Disaster Declarations Summary, a dataset that summarizes all federally declared disasters. The record goes back to the first federal dissaster declaration in 1953, a tornado on May 2nd in Georgia. The Disaster Declarations Summary is raw unedited data much of which was manually entered from highly variable historical records. This analysis will tackle the unique structure and idiosynchracies of the dataset to present an analysis disaster durations for various categories of Federal disasters. Additionally, regionally differences in disaster duration will be explored.
Data Preparation
Loading libraries used
Loading the Disaster Declaration Dataset to an R data.frame
URL <- "https://raw.githubusercontent.com/SmilodonCub/DATA606/master/DisasterDeclarationsSummaries1.csv"
#subset the data for features relevant to this analysistime duration between declarationDate and disasterCloseOutDate
#calculate the
disaster_df <- read_csv( URL ) ## Parsed with column specification:
## cols(
## disasterNumber = col_double(),
## ihProgramDeclared = col_double(),
## iaProgramDeclared = col_double(),
## paProgramDeclared = col_double(),
## hmProgramDeclared = col_double(),
## state = col_character(),
## declarationDate = col_datetime(format = ""),
## fyDeclared = col_double(),
## disasterType = col_character(),
## incidentType = col_character(),
## title = col_character(),
## incidentBeginDate = col_datetime(format = ""),
## incidentEndDate = col_datetime(format = ""),
## disasterCloseOutDate = col_datetime(format = ""),
## declaredCountyArea = col_character(),
## placeCode = col_double(),
## hash = col_character(),
## lastRefresh = col_datetime(format = ""),
## id = col_character()
## )disaster_df <- disaster_df%>%
select( disasterNumber, disasterType, declarationDate, disasterCloseOutDate,
incidentType, ihProgramDeclared, iaProgramDeclared, paProgramDeclared,
hmProgramDeclared,incidentBeginDate, incidentEndDate, state, declaredCountyArea,
placeCode) %>%
mutate( interval = declarationDate %--% disasterCloseOutDate,
duration_secs = int_length( interval ))
head( disaster_df )## # A tibble: 6 x 16
## disasterNumber disasterType declarationDate disasterCloseOutDa…
## <dbl> <chr> <dttm> <dttm>
## 1 1 DR 1953-05-02 00:00:00 1954-06-01 00:00:00
## 2 2 DR 1953-05-15 00:00:00 1958-01-01 00:00:00
## 3 3 DR 1953-05-29 00:00:00 1960-02-01 00:00:00
## 4 6 DR 1953-06-09 00:00:00 1956-03-30 00:00:00
## 5 4 DR 1953-06-02 00:00:00 1956-02-01 00:00:00
## 6 5 DR 1953-06-06 00:00:00 1955-12-01 00:00:00
## # … with 12 more variables: incidentType <chr>, ihProgramDeclared <dbl>,
## # iaProgramDeclared <dbl>, paProgramDeclared <dbl>, hmProgramDeclared <dbl>,
## # incidentBeginDate <dttm>, incidentEndDate <dttm>, state <chr>,
## # declaredCountyArea <chr>, placeCode <dbl>, interval <Interval>,
## # duration_secs <dbl>Research question
You should phrase your research question in a way that matches up with the scope of inference your dataset allows for.
Has Disaster Closeout time changed significantly in the time that FEMA has been issuing disaster declarations? Is the change consistent across disaster type? Geographical region?
Cases
What are the cases, and how many are there?
Each case represents a FEMA declared disaster event. There are nrow( disaster_df ) events documented in this data set.
Data collection
Describe the method of data collection.
Data is collected by the Federal Emergency Management Agency (FEMA) as part of the OpenFEMA Dataset (OpenFEMA) in an effort to make data freely available in machine readable formats
Type of study
What type of study is this (observational/experiment)?
This is an observational study.
Data Source
If you collected the data, state self-collected. If not, provide a citation/link.
Data is collected by OpenFEMA and is available online here: OpenFEMA Dataset: Disaster Declarations Summaries - V2 For this project, data was uploaded to the authors github to facilitate access.
Response
What is the response variable, and what type is it (numerical/categorical)?
The response variable is the disaster duration which is to be calculated from the difference between the ‘incidentBeginDate’ and ‘incidentEndDate’ features and is a numerical variable.
Explanatory
What is the explanatory variable, and what type is it (numerical/categorical)?
The explanatory variables are categorical: disaster type (‘incidentType’) and geographical region (fipsStateCode & fipsCountyCode).
Relevant summary statistics
Provide summary statistics relevant to your research question. For example, if you’re comparing means across groups provide means, SDs, sample sizes of each group. This step requires the use of R, hence a code chunk is provided below. Insert more code chunks as needed.
Summary Statistics for Disaster Durations
#format date interval
disaster_df$duration_days <- as.numeric(as.character(day(seconds_to_period(disaster_df$duration_secs))))
print('Summary Statistics given in Days')## [1] "Summary Statistics given in Days"## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 0 1312 2500 2592 3563 12107 10017Summary Statistics for Duration (Days) by Disaster Type
groupedbyDType <- disaster_df %>%
group_by( disasterType ) %>%
summarise( count = n(),
mean = mean(duration_days, na.rm = TRUE),
median = median(duration_days, na.rm = TRUE),
IQR = IQR(duration_days, na.rm = TRUE)
)
groupedbyDType## # A tibble: 4 x 5
## disasterType count mean median IQR
## <chr> <int> <dbl> <dbl> <dbl>
## 1 DR 36498 3125. 3009 1815
## 2 EM 13097 1506. 1108 1203
## 3 FM 1031 1123. 1140 666
## 4 FS 390 1248. 1058 661.Boxplot of Disaster Types Summary Stats
typesD <- c('Major Disaster', 'Emergency Declaration', 'Fire Management', 'Fire Suppression')
ggplot(groupedbyDType, aes(x = as.factor(disasterType))) +
geom_boxplot(aes(
lower = median - IQR,
upper = median + IQR,
middle = median,
ymin = median - 3*IQR,
ymax = median + 3*IQR),
stat = "identity") +
ggtitle('Disaster Type Duration Comparison') +
xlab('Disaster Type') +
ylab('Duration (days)') +
scale_x_discrete(labels = typesD)