distrExIntegrate {distrEx}  R Documentation 
Numerical integration via integrate
. In case integrate
fails a GaussLegendre quadrature is performed.
distrExIntegrate(f, lower, upper, subdivisions = 100,
rel.tol = .Machine$double.eps^0.25,
abs.tol = rel.tol, stop.on.error = TRUE,
distr, order, ..., diagnostic = FALSE)
showDiagnostic(x, what, withNonShows = FALSE, ...)
getDiagnostic(x, what, reorganized=TRUE)
## S3 method for class 'DiagnosticClass'
print(x, what, withNonShows = FALSE, xname, ...)
f 
an R function taking a numeric first argument and returning a numeric vector of the same length. Returning a nonfinite element will generate an error. 
lower 
lower limit of integration. Can be 
upper 
upper limit of integration. Can be 
subdivisions 
the maximum number of subintervals. 
rel.tol 
relative accuracy requested. 
abs.tol 
absolute accuracy requested. 
stop.on.error 
logical. If 
distr 
object of class 
order 
order of GaussLegendre quadrature. 
diagnostic 
logical; if 
... 
In case of integrators: additional arguments to be passed to 
x 
the item for which the diagnostic is to be shown. 
what 
a character vector with all the diagnostic items to be selected/shown. If empty or missing all items are selected/shown. 
withNonShows 
internally we distinguish items which are easily
printed (first kind) (numeric, logical, character)
and more difficult ones (second kind), e.g., calls, functions, lists. The
distinction is made according to the list item name. If

xname 
an optional name for the diagnostic object to be shown. 
reorganized 
should the diagnostic information be reorganized
(using internal function 
distrExIntegrate
calls integrate
. In case integrate
returns an error a GaussLegendre integration is performed using
GLIntegrate
. If lower
or (and) upper
are infinite
the GLIntegrateTruncQuantile
, respectively the
1GLIntegrateTruncQuantile
quantile of distr
is used
instead.
distrExIntegrate
is called from many places in the distr and robast
families of packages. At every such instance, diagnostic information can
be collected (setting a corresponding argument diagnostic
to TRUE
in the calling function. This diagnostic information is originally stored in
a tree like list structure of S3 class DiagnosticClass
which is then
attached as attribute diagnostic
to the respective object. It can be
inspected and accessed through showDiagnostic
and getDiagnostic
.
More specifically, for any object with attribute diagnostic
,
showDiagnostic
shows the diagnostic collected during integration, and
getDiagnostic
returns the diagnostic collected during integration.
To this end, print.DiagnosticClass
is an S3 method for print
for objects of S3 class DiagnosticClass
.
The value of distrExIntegrate
is a numeric approximation of the integral.
If argument diagnostic==TRUE
in distrExIntegrate
, the return
value has an attribute diagnostic
of S3 class DiagnosticClass
containing diagnostic information on the integration.
showDiagnostic
, getDiagnostic
, print.DiagnosticClass
all
return (invisibly) a list with the selected items, reorganized by internal
function .reorganizeDiagnosticList
, respectively, in case
of argument reorganized==FALSE
, getDiagnostic
returns (invisibly)
the diagnostic information as is.
Matthias Kohl Matthias.Kohl@stamats.de
Based on QUADPACK routines dqags
and dqagi
by
R. Piessens and E. deDonckerKapenga, available from Netlib.
R. Piessens, E. deDonckerKapenga, C. Uberhuber, D. Kahaner (1983) Quadpack: a Subroutine Package for Automatic Integration. Springer Verlag.
W.H. Press, S.A. Teukolsky, W.T. Vetterling, B.P. Flannery (1992) Numerical Recipies in C. The Art of Scientific Computing. Second Edition. Cambridge University Press.
integrate
, GLIntegrate
,
distrExOptions
fkt < function(x){x*dchisq(x+1, df = 1)}
integrate(fkt, lower = 1, upper = 3)
GLIntegrate(fkt, lower = 1, upper = 3)
try(integrate(fkt, lower = 1, upper = 5))
distrExIntegrate(fkt, lower = 1, upper = 5)