This is a method for the generic command `with`

for an object of class `"fv"`

(function value table).

An object of class `"fv"`

is a convenient way of storing and
plotting several different estimates of the same function. It is
effectively a data frame with extra attributes.
See `fv.object`

for further explanation.

This command makes it possible to perform computations that involve
different estimates of the same function. For example we use it to compute
the arithmetic difference between two different edge-corrected
estimates of the \(K\) function of a point pattern.

The argument `expr`

should be an R language expression. The expression
may involve

the name of any column in `data`

, referring to
one of the estimates of the function;

the symbol `.`

which stands for all
the available estimates of the function;

the symbol `.y`

which stands for the recommended
estimate of the function (in an `"fv"`

object, one of the
estimates is always identified as the recommended estimate);

the symbol `.x`

which stands for the argument of the function;

global constants or functions.

See the Examples.
The expression should be capable of handling
vectors and matrices.

The interpretation of the argument `fun`

is as follows:

If `fun=FALSE`

, the result of evaluating the expression
`expr`

will be returned as a numeric vector, matrix or
data frame.

If `fun=TRUE`

, then the result of evaluating `expr`

will be interpreted as containing the values of a new function.
The return value will be an object of class `"fv"`

.
(This can only happen if the result has the right dimensions.)

The default is `fun=TRUE`

if the result of evaluating
`expr`

has more than one column, and `fun=FALSE`

otherwise.

To perform calculations involving *several* objects of
class `"fv"`

, use `eval.fv`

.