CODEWITHSUNDEEP
r
AS1989
AS1989

Reputation: 1

creating 2x2 table in R with the values of cells being a 3rd variable, not a count

I want to create a table in R from a larger data set. The table would be 2x2, however the values within each cell would be the value of a third variable already in the dataset. Specifically, I am looking at the number of prescriptions for a class of medication by year. Thus one side of the the table would be drug-class (either A or B), the other year (2014-2018), and each cell would represent number of prescriptions.

In the dataset, each row contains statistics for a given medication in a given year. Each row is not an individual prescription/patient. There is a column for number of prescriptions. Each. All the summarize functions are giving me counts which is not what I am looking for.

Ultimately I would like to compare the proportion of patients in each medication class by year with a chi-square test.

Upvotes: 0

Views: 784

Answers (1)

dcarlson
dcarlson

Reputation: 11076

Make up some data:

set.seed(42)
Diuretic <- sample(c("yes", "no"), 100, replace=TRUE)
Year <- sample(c(2014, 2015, 2016), 100, replace=TRUE)
Beneficiaries <- round(rnorm(100, 35, 5))
dta <- data.frame(Diuretic, Year, Beneficiaries)

Now use xtabs:

(dta.tbl <- xtabs(Beneficiaries~Diuretic+Year, dta)
#         Year
# Diuretic 2014 2015 2016
#      no   741  888  295
#      yes  448  649  429

Add totals:

addmargins(dta.tbl)
#         Year
# Diuretic 2014 2015 2016  Sum
#      no   741  888  295 1924
#      yes  448  649  429 1526
#      Sum 1189 1537  724 3450

Upvotes: 1

Related Questions