Cumulative Distribution Function

DESCRIPTION:

Empirical cumulative distribution function.

USAGE:

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.

SEE ALSO:

, , , .

EXAMPLES:

boot <- bootstrap(1:9, mean) 
cdf(boot, 3:7) 
cdf(boot$replicates, 3:7) 
cdf(boot, boot$observed)