Skip to contents

Distribution function of a set-valued discrete variable

Source: R/interval.R
tda_sddf.Rd

The set-valued counterpart of tda_iddf: each case's observation is a set of possible categories (not necessarily a single value, and not necessarily an interval), and the distribution function is only identified up to bounds consistent with every case's own set.

Usage

tda_sddf(
  sets,
  categories = NULL,
  self_consistent = FALSE,
  control = NULL,
  options = list(),
  dir = tempfile("tda"),
  ...
)

Arguments

sets

a list, one element per case, each the vector of category labels that case's set contains – list(1, 2, 3, c(1, 2), c(2, 3)) for the five cases in TDA's manual example (\(o_1 = \{1\}\), ..., \(o_5 = \{2, 3\}\)). Converted internally to the indicator columns sddf itself needs (one 0/1 column per category, 1 if that case's set contains it), the shape TDA's manual example uses. A data frame or matrix of 0/1 indicator columns (one column per category, exactly sddf's shape – what a plain read.table() import of a file like TDA's id1.dat already is) is accepted directly too, with no conversion needed.

categories

the full list of possible categories; by default, the sorted union of everything appearing in sets.

self_consistent

also compute the self-consistent distribution (opt=2), not just the min/max/mean bounds (opt=1, default) – sddf's opt=.

control

convergence settings (mxit=/tolf via tda_control), only meaningful with self_consistent = TRUE.

options

a named list of further TDA options, passed through.

dir

working directory.

...

passed to tda_run.

Value

An object carrying a table of categories and the distribution function's bounds and mean at each.

See also

Other interval-valued data: tda_bounds(), tda_idf(), tda_ilsreg(), tda_imean(), tda_imreg(), tda_inpreg(), tda_ivar1(), tda_ivreg()

Examples

# TDA's manual example: o1={1}, o2={2}, o3={3}, o4={1,2}, o5={2,3}
fit <- tda_sddf(list(1, 2, 3, c(1, 2), c(2, 3)))
fit$table
#>   category lower upper mean_df
#> 1        1   0.2   0.4     0.3
#> 2        2   0.2   0.6     0.4
#> 3        3   0.2   0.4     0.3

# the same data, already read in as indicator columns (e.g. from
# id1.dat via read.table()) -- no conversion needed
id1 <- data.frame(X1 = c(1, 0, 0, 1, 0), X2 = c(0, 1, 0, 1, 1),
                  X3 = c(0, 0, 1, 0, 1))
tda_sddf(id1)$table
#>   category lower upper mean_df
#> 1       X1   0.2   0.4     0.3
#> 2       X2   0.2   0.6     0.4
#> 3       X3   0.2   0.4     0.3