cdf(x, q, weights = x$weights, na.rm=F, normalize=T)
REQUIRED ARGUMENTS:
x
a
resamp object; the
replicates component is extracted and passed
to the usual
default method.
q
numerical, values at which to compute the distribution function.
This may also be a matrix with
p columns (where
x$replicates has
B rows and
p columns).
OPTIONAL ARGUMENTS:
weights
a vector of length
B, containing non-negative weights
(probabilities)
to define a distribution function with unequal jumps.
The weights are normalized to sum to 1.
na.rm
if
TRUE, missing values (
NA) in
x or
weights cause
the corresponding observations to be removed.
If
FALSE, missing values in
x and
weights are an error.
normalize
if
FALSE, then return the number (rather than proportion)
of the
x values less than or equal to each q,
or the sum of the corresponding unnormalized weights.
VALUE:
matrix with
length(q) rows and
p columns, column
j
containing F(
q), where F is
the empirical distribution function defined by column
j of
x$replicates.
Missing values in
q cause missing values in the result.