The goal of tothewonder is to make it easy to access CDC
WONDER mortality statistics. This package was
created to protest the REPUBLICAN and NRA led efforts to suppress
research into the underlying causes of gun violence.
devtools::install_github("yukatapangolin/tothewonder", ref = "master")or download the source zip file from github and unzip
library(devtools)
source <- devtools:::source_pkg("C:/path/tothewonder-master")
install(source)To use this package you’ll need to start a session at the CDC WONDER website and agree to their data use restrictions
Depending on which database you want to access you can use one of the following functions to start a session:
session_ucd99() - Underlying Cause of Death,
1999-2020
session_mcd_provisional() - Provisional Mortality Statistics, 2018
through Last Month
session_mcd_final18() - Multiple Cause of Death, 2018-2021, Single
Race
After agreeing to abide by the data use restrictions the functions will return a string with the CDC WONDER session URL.
Note that it’s possible this package violates the terms of use of the
CDC WONDER website since they have a public
API that’s
restricted to only return data at the national level and tothewonder
allows you retrieve sub-national data, but the web
application already lets you retrieve
sub-national and it’s probably no big deal to use this package
However, in keeping with the vital statistics policy for public data sharing, only national data are available for query by the API. Queries for mortality and births statistics from the National Vital Statistics System cannot limit or group results by any location field, such as Region, Division, State or County, or Urbanization (urbanization categories map to specific geographic counties). For example, in the D76 online database for Detailed Mortality 1999-2013, the location fields are D76.V9, D76.V10 and D76.V27, and the urbanization fields are D76.V11 and D76.V19. These ’sub-national" data fields cannot be grouped by or limited via the API, although these fields are available in the web application. https://wonder.cdc.gov/wonder/help/WONDER-API.html
It’s the opinion of the tothewonder author that these restrictions
make the official API useless. Additionally, since the web application
already permits querying at the sub-national level the author sees no
problem in automating this type of query as long as users follow the
data use restrictions. The author
of this package furthermore believes that these limitations were put in
place thanks to RUSSIAN PUPPET PRESIDENT DONALD TRUMP to prevent
researchers from investigating GUN VIOLENCE.
Sanctions for Violating Rules:
Researchers who violate the terms of the data use restrictions will lose access to WONDER and their sponsors and institutions will be notified. Researchers who are suspected of violating the rules may be prevented from using WONDER until an investigation can be completed. Deliberately making a false statement in any matter within the jurisdiction of any department or agency of the Federal government violates 18 USC 1001 and is punishable by a fine of up to $10,000 or up to 5 years in prison, or both.
Depending on which database you want to access you can use one of the following functions to download your data:
ucd99() - Underlying Cause of Death,
1999-2020
mcd_provisional() - Provisional Mortality Statistics, 2018 through
Last Month
mcd_final18() - Multiple Cause of Death, 2018-2021, Single
Race
library(tothewonder)
## Start a UCD (https://wonder.cdc.gov/ucd-icd10.html) session and
## agree to the data use restrictions
url_ucd99 <- session_ucd99()
## Download monthly non-Hispanic Black firearm deaths data for
## Missouri (FIPS code 29). This would include homicides, suicides, accidents
## and legal intervertions/operations of war. Singel Age Years 15-55 only
df <- ucd99(wonder_url = url_ucd99,
group_by_1 = "Year",
group_by_2 = "Month",
group_by_3 = "None",
group_by_4 = "None",
show_totals = FALSE,
show_confidence_interval = FALSE,
show_standard_error = FALSE,
age = 15:55,
period = 1999:2020,
residence_urbanization_year = "2013",
residence_urbanization = "All Categories",
residence_fips = "29", # Missouri has FIPS 29
weekday = c("All Weekdays"),
autopsy = c("All Values"),
place_of_death = c("All Places"),
gender = c("All"),
hispanic_origin = "Not Hispanic or Latino",
race = "Black or African American",
ucd_option = "Injury Intent and Mechanism",
ucd_injury_intent = "All Causes of Death",
ucd_injury_mechanism = "Firearm"
)
## Start a session with the Provisional Mortality by Multiple Cause of Death db
url_provisional <- session_mcd_provisional()
## Download weekly data where the underlying cause of death was homicide, and
## where the victim was African American
## Only include the "15-24","25-34", and "35-44" ten year age groups
df <- mcd_provisional(wonder_url = url_provisional,
group_by_1 = "MMWR Week",
group_by_2 = "None",
group_by_3 = "None",
group_by_4 = "None",
show_totals = FALSE,
show_confidence_interval = FALSE,
show_standard_error = FALSE,
age = c("15-24","25-34", "35-44"),
period_option = "MMWR",
period = 2018:2022,
residence_fips = "All",
residence_urbanization_year = c("2013"),
residence_urbanization = "All Categories",
occurrence_fips = "All",
occurrence_urbanization_year = c("2013"),
occurrence_urbanization = "All Categories",
autopsy = c("All Values"),
place_of_death = c("All Places"),
gender = c("All Genders"),
hispanic_origin = "Not Hispanic or Latino",
race_option = "Single Race 6",
race = "Black or African American",
ucd_option = "Injury Intent and Mechanism",
ucd_injury_intent = "Homicide",
ucd_injury_mechanism = "All Causes of Death",
mcd_option = "MCD - ICD-10 Codes",
mcd_icd_codes = "All",
mcd_icd_codes_and = "All")You can also save queries to a unique link that leads to the WONDER
website by using the save parameter
save_session <- session_ucd99()
ucd99(wonder_url = save_session,
save = TRUE,
group_by_1 = "Year",
group_by_2 = "Month",
group_by_3 = "None",
group_by_4 = "None",
show_totals = FALSE,
show_confidence_interval = FALSE,
show_standard_error = FALSE,
age = 15:55,
period = 1999:2020,
residence_urbanization_year = "2013",
residence_urbanization = "All Categories",
residence_fips = "29", # Missouri has FIPS 29
weekday = c("All Weekdays"),
autopsy = c("All Values"),
place_of_death = c("All Places"),
gender = c("All"),
hispanic_origin = "Not Hispanic or Latino",
race = "Black or African American",
ucd_option = "Injury Intent and Mechanism",
ucd_injury_intent = "All Causes of Death",
ucd_injury_mechanism = "Firearm"
)Which should output something like:
“https://wonder.cdc.gov/controller/saved/D76/D296F514”
(each query you save will generate a unique link even if all the parameters are the same)
Firearm homicide rate by county level presidential vote in 2020
library(tothewonder)
library(tidyverse)
## from https://github.com/tonmcg/US_County_Level_Election_Results_08-20
url <- paste0("https://raw.githubusercontent.com/yukatapangolin/",
"US_County_Level_Election_Results_08-20/",
"585cb48ccc17fef3640db33eef952cf613c67e95/",
"2020_US_County_Level_Presidential_Results.csv")
trump_2020 <- read.csv(url,
colClasses = c("character", "character",
"character", "integer",
"integer", "integer",
"integer", "numeric",
"numeric", "numeric"))
trump_2020$winner <- ifelse(trump_2020$per_gop > trump_2020$per_dem,
"Trump", "Biden")
trump_2020 <- subset(trump_2020, per_gop >= .6 | per_dem >= .6)
trump_2020 <- subset(trump_2020, state_name != "Alaska")
## Shannon County (46113) became Oglala Lakota County (46102) in 2015 and
## WONDER doesn't know about the new FIPS code
## https://en.wikipedia.org/wiki/Oglala_Lakota_County,_South_Dakota
trump_2020$county_fips[which(trump_2020$county_fips == 46102)] <- 46113
url_mcd18 <- session_mcd_final18(I_Agree = TRUE)
deaths <- map_dfr( unique(trump_2020$winner), .f = \(x) {
counties <- unique(subset(trump_2020, winner == x))$county_fips
## Parameters to functions use the same defaults
## as the ones available at CDC WONDER and you
## can omit them
df <- mcd_final18(wonder_url = url_mcd18,
save = FALSE,
group_by_1 = "Year",
period = 2018:2021,
residence_fips = counties,
gender = "Male",
ucd_option = "Injury Intent and Mechanism",
ucd_injury_intent = "Homicide",
ucd_injury_mechanism = "Firearm")
df$winner <- x
return(df)
})
deaths |>
mutate(deaths = as.numeric(Deaths)) |>
ggplot(aes(Year.Code, Crude.Rate, group = winner, color = winner))
geom_line(linewidth = .9)
labs(title = "Male Firearm Homicide Rate by 2020 Presidential County Vote",
subtitle = paste0("Counties that voted for Trump in 2020",
" experienced smaller homicide increases."),
caption = paste0("Source: CDC WONDER. ",
"2018-2021 Final Multiple Cause of Death Files"))
expand_limits(y = 0)
scale_color_manual(
values = c("#244999",
"#d22532"),
labels = c("Biden won with 60% of the vote",
"Trump won with 60% of the vote")
)
ylab("Rate per 100,000")
xlab("Year")
-
Bump the version using the bump2version command (pip install –upgrade bump2version).
-
Update package data by running
Rscript data-raw/DATASET.R -
Run
Rscript -e "devtools::document(roclets = c('rd', 'collate', 'namespace'))"to update documentation -
Run
find . -type f -exec sedto correct paths -
Run
Rscript -e 'rmarkdown::render("README.Rmd")'to knit README.Rmd -
Update the NEWS.md file with any changes.
-
Tag the release
git tag v0.1.0 && git push origin --tags -
Checkout development branch