The names of the functions to apply to numeric data.
xname
A name for the object x, mostly where this would be
a very long string describing its structure (e.g. if it was extracted
by name from a data frame).
horizontal
Whether to display the results of describe.factor
across the page (TRUE) or down the page (FALSE).
Details
describe displays a table of descriptive statistics for numeric,
factor and logical variables in the object x. The summary measures
for numeric variables can easily be altered with the argument
num.desc. Pass a character vector with the names of the
desired summary measures and these will be displayed at the top of
the numeric block with their results beneath them. If quantiles are
desired, the user will have to write wrapper functions that call
quantile with the appropriate type or probabilities and use
the names of the wrapper functions in num.desc. Remember that
any function called by describe must have an na.rm
argument.
Percentages are now always displayed and returned in the tables for
factor and logical variables.
Value
A list with three components:
Numeric
A list of the values returned from describe.numeric
for each column described.
Factor
A list of the tables for each column described.
Logical
A list of the tables for each column described.
Author(s)
Jim Lemon
See Also
Mode, valid.n, describe.numeric,
describe.factor
Examples
sample.df<-data.frame(sex=sample(c("MALE","FEMALE"),100,TRUE),
income=(rnorm(100)+2.5)*20000,suburb=factor(sample(1:4,100,TRUE)))
# include the maximum values
describe(sample.df,num.desc=c("mean","median","max","var","sd","valid.n"))
# make up a function
roughness<-function(x,na.rm=TRUE) {
if(na.rm) x<-x[!is.na(x)]
xspan<-diff(range(x))
return(mean(abs(diff(x))/xspan,na.rm=TRUE))
}
# now use it
describe(sample.df$income,num.desc="roughness",xname="income")