This command is the counterpart of `envelope`

when the function `fun`

that is evaluated on each simulated point pattern
will return an object of class `"fasp"`

representing an array of
summary functions.

Simulated point patterns are generated according to the
rules described for `envelope`

. In brief,
if `X`

is a point pattern, the algorithm generates
simulated point patterns of the same kind, according to complete
spatial randomness. If `X`

is a fitted model, the algorithm
generates simulated point patterns according to this model.

For each simulated point pattern `Y`

, the function `fun`

is invoked. The result `Z <- fun(Y, ...)`

should be an object of
class `"fasp"`

representing an array of summary functions.
The dimensions of the array `Z`

should be the same
for each simulated pattern `Y`

.

This algorithm finds the simulation envelope of the summary functions
in each cell of the array.