TDA's ineq: for each variable, the (weighted) minimum, maximum,
mean and standard deviation, the coefficient of variation (standard
deviation over mean), and the Gini coefficient – the mean absolute
difference between every pair of values, relative to twice the mean,
the standard concentration measure for incomes and similar
quantities: 0 when every value is equal, approaching 1 as the total
concentrates in one case.
Value
An object carrying table: one row per variable, with
cases (nonnegative values only), minimum,
maximum, mean, sd, vcoeff and
gini.
Details
ineq treats every negative value as missing and drops
it – the manual (section 6.4.2) is explicit that only nonnegative
values enter, so cases in the result can be smaller than the
number of rows given, and a variable that legitimately runs negative
is not summarised correctly by this command. Case weights given via
a prior cwt command (options, or the weights
arguments of the model functions, set one) weight every measure.
See also
Other descriptive statistics:
tda_brr(),
tda_cov(),
tda_dstat(),
tda_freq(),
tda_independence(),
tda_loglin(),
tda_quant(),
tda_rcorr(),
tda_segr(),
tda_subm()
Examples
set.seed(29)
d <- data.frame(income = round(rlnorm(80, meanlog = 8, sdlog = 0.6)))
tda_ineq(d)$table
#> index cases minimum maximum mean sd vcoeff gini
#> 1 income 80 692 11148 3293.725 2212.713 0.6717966 0.3429973