Is there any easy way to make a such a table for discrete in R: 1. the value in table is percentage for each row. 2. divide by more than two groups. For example, the data is
Success Gender Level
1 M High
1 M Low
1 F Med
0 M Low
0 M Med
0 F High
The desired table looks like this
Success=1 Success=0
Level=High Level=Med Level=Low Level=High Level=Med Level=Low
Gender=F 0 0.5 0 0.5 0 0
Gender=M 0.25 0 0.25 0 0.25 0.25
You can use ftable() with prop.table(). Specifying row.vars as the second column will produce a table that looks like your desired table (in a slightly different order).
prop.table(ftable(df, row.vars = 2), margin = 1)
# Success 0 1
# Level High Low Med High Low Med
# Gender
# F 0.50 0.00 0.00 0.00 0.00 0.50
# M 0.00 0.25 0.25 0.25 0.25 0.00
For the exact desired table, you can refactor columns to change the order of the levels.
df2 <- transform(
df,
Level = factor(Level, levels = c("High", "Med", "Low")),
Success = factor(Success, levels = 1:0)
)
prop.table(ftable(df2, row.vars = 2), margin = 1)
# Success 1 0
# Level High Med Low High Med Low
# Gender
# F 0.00 0.50 0.00 0.50 0.00 0.00
# M 0.25 0.00 0.25 0.00 0.25 0.25
Data:
df <- structure(list(Success = c(1L, 1L, 1L, 0L, 0L, 0L), Gender = structure(c(2L,
2L, 1L, 2L, 2L, 1L), .Label = c("F", "M"), class = "factor"),
Level = structure(c(1L, 2L, 3L, 2L, 3L, 1L), .Label = c("High",
"Low", "Med"), class = "factor")), .Names = c("Success",
"Gender", "Level"), class = "data.frame", row.names = c(NA, -6L
))