This project is a work-in-progress. It works in some cases but may not work in many others (and may not be flexible enough for some users), and some extra code exists for functions that are not yet operational. This program has only been tested in Windows 10.
Before you can use the get_data function, you need to:
- Ensure that you have Python3 installed. You will need to locate your python.exe file. In Windows, you can run the following to figure out where that is:
where python- Install the cdcwonderpy package for Python3
pip install cdcwonderpy- Install PythonInR in R
install.packages('PythonInR')
library(PythonInR)
- Install this package
library(devtools)
devtools::install_github('tlcaputi/wondeR')
The first step is to collect data. This implements the wonderpy package for Python3 within R. For demonstration, we pull the data for all drug overdoses, which is all deaths that include at least one ICD10 code in T36-T50 AND at least one ICD 10 code in X40-X44, X60-X64, X85, or Y10-Y14.
library(wondeR)
df <- get_data(
ROOTPATH = "/path/to/root", # this should include an "input" and "output" subdirectory
RUNNAME = "sex_race_overdose", # Name of the run
replace = T, # Default True, if False then this will not collect data
pypath = "/path/to/python.exe", # Path to your python3.exe file (see above)
## Collect data on deaths that include these ICD10 codes
MCD1 = c("T36-T50"), ## Any of these codes...
MCD2 = c("X40-X44", "X60-X64", "X85", "Y10-Y14"), ## AND any of these codes. Defaults to c(NA), which is All death codes
## Segment the data based upon these variables
by_vars = c("state", "sex", "race") # can be sex, age, race, hispanic, state
)
Note that you can use code ranges in the MCD1 or MCD2 arguments.
If you want to collect the data yourself or using the wonderpy package within Python, you can make the replace argument False and ensure the data file matches the format of RUNNAME + "_pull.csv" in your input subdirectory. In this case, you'd want to make sure that "ROOTPATH/input//opioids_pull.csv" exists. You should still run the data through get_data before moving onto future functions.
Combine this data with the plot_grid function to create line plots. Here, we will look at how drug overdose deaths have changed by sex and by race (grid_vars), comparing states that expanded Medicaid under the Affordable Care Act against those that did not (groups and group_conditions).
out <- plot_grid(
df, # Data frame from get_data
## Should be same as above
ROOTPATH = "C:/Users/tcapu/Google Drive/modules/wondeR/READMEcode",
RUNNAME = "sex_race_overdose",
## Convenience arguments
minwonderyear = 1999, # Min year that WONDER collects
maxwonderyear = 2018, # Max year that WONDER collects
## Multiple Lines within the Same Plot
# Create the names for the groups that you want
groups = c(
"No Expansion Before 2019",
"Expanded in 2014",
"Expanded Between 2015 and 2018"
),
# Write the conditions for those groups as strings
group_conditions = c(
"is.na(aca_date) | year(aca_date) > 2018",
"year(aca_date) == 2014",
"T"
),
# Legend Title for the Groups
group_title = "Medicaid",
# Will add n= values to the group labels
include_n = T,
# If True, will only include data from state-years with data for all groups
listwise = F,
## Multiple plots in a grid
# Segment data by x and y
grid_vars = c("sex", "race"),
## Plot Arguments
xlab = "Year",
ylab = "Crude Rate\n(Per 100k Deaths)",
colpalette = "Dark2", # ggplot theme
vline = T, # If logical, no vertical line. If numeric/date, a dotted vertical line
## Plot Saving arguments
save = T, # If True, a plot will be saved
out_fn = T, # If logical, will be named after RUNNAME
width = 6, # width in inches
height = 4, # height in inches
## Do you want to return the data or just the plot?
include_data = T
)
If you set include_data = T, the resulting plot (a ggplot) will be the first item in the resulting list.
p <- out[[1]]
ggsave("./output/Fig1.png", p, width = 10, height = 7)
