Least squares with an interval-valued response

The response is interval-valued, the regressor a point value – unlike tda_ivreg, where both sides are intervals. Both the slope and the intercept come back as bounds; see tda_bounds.

Usage

tda_ilsreg(
  formula,
  data,
  yl,
  censor,
  options = list(),
  dir = tempfile("tda"),
  ...
)

Arguments

formula

a formula using iv() for the interval-valued response, e.g. iv(ylo, yhi) ~ x.

data

a data frame.

yl, censor

the response and censoring variables TDA requires, as names or vectors.

options

a named list of further TDA options, passed through.

dir

working directory.

...

passed to tda_run.

Value

An object of class tda_fit.

See also

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

Examples

set.seed(2)
d <- data.frame(x = round(rnorm(30), 2))
d$ylo <- 2 + 0.5 * d$x + rnorm(30, sd = 0.2)
d$yhi <- d$ylo + round(runif(30, 0.3, 1), 2)
tda_ilsreg(iv(ylo, yhi) ~ x, d)
#> Call: tda_ilsreg(iv(ylo, yhi) ~ x, d)
#> 
#> Cases: 30 
#> 
#> Bounds on the slope: [0.2378364, 0.6916871]
#> Bounds on the intercept: [2.270285, 2.36743]