For a one sample problem, compares the empirical distribution function (edf)
of the sample with a hypothesized cumulative distribution function.
For a two sample problem,
compares the edfs for the two samples. This graphical comparison
is often useful
before performing the
Kolmogorov-Smirnov test (function
ks.gof).
USAGE:
cdf.compare(x, y = NULL, distribution = "normal", ...)
REQUIRED ARGUMENTS:
x
numeric vector.
NAs and
Infs are allowed but will be
removed.
OPTIONAL ARGUMENTS:
y
numeric vector.
NAs and
Infs are allowed but will be removed.
distribution
character string that specifies the hypothesized distribution in the
one sample test. For two samples, i.e. when
y is specified, this
argument is ignored.
distribution can be one of:
"normal", "beta", "cauchy", "chisquare", "exponential", "f", "gamma",
"lognormal",
"logistic", "t", "uniform", "weibull", "binomial", "geometric",
"hypergeometric",
"negbinomial", "poisson", or "wilcoxon".
You need only supply the first characters that
uniquely specify the distribution name. For example, "logn"
and "logi" uniquely specify the lognormal and logistic
distributions.
...
For one sample, parameter arguments for
the S-PLUS function that generates p-values for
the hypothesized distribution.
SIDE EFFECTS:
Produces a plot of the two compared cdfs on the current graphics device.
SEE ALSO:
,
,
(to create QQ plots),
,
,
..
EXAMPLES:
# one sample
z <- rnorm(100)
cdf.compare(z,dist="normal") #compare with a normal distn.
cdf.compare(z,dist="chisquare",df=2) #compare with a chisquare distn.
# two sample
x <- rnorm(25)
y <- rexp(100)
cdf.compare(x,y)