Converts an object of class `"fv"`

to an R language function.

```
# S3 method for fv
as.function(x, ..., value=".y", extrapolate=FALSE)
```

A `function(r)`

or `function(r,what)`

where `r`

is the name of the original argument of the function table.

- x
Object of class

`"fv"`

or`"rhohat"`

.- ...
Ignored.

- value
Optional. Character string or character vector selecting one or more of the columns of

`x`

for use as the function value. See Details.- extrapolate
Logical, indicating whether to extrapolate the function outside the domain of

`x`

. See Details.

Adrian Baddeley Adrian.Baddeley@curtin.edu.au and Rolf Turner r.turner@auckland.ac.nz

A function value table (object of class `"fv"`

)
is a convenient way of storing and plotting
several different estimates of the same function.
Objects of this class are returned
by many commands in spatstat,
such as `Kest`

,
which returns an estimate of Ripley's \(K\)-function
for a point pattern dataset.

Sometimes it is useful to convert the function value table
to a `function`

in the R language. This is done by
`as.function.fv`

. It converts an object `x`

of class `"fv"`

to an R function `f`

.

If `f <- as.function(x)`

then `f`

is an R function
that accepts a numeric argument and returns a corresponding value
for the summary function by linear interpolation between the values
in the table `x`

.

Argument values lying outside the range of the table
yield an `NA`

value (if `extrapolate=FALSE`

)
or the function value at the nearest endpoint of the range
(if `extrapolate = TRUE`

).
To apply different rules to the left and right extremes,
use `extrapolate=c(TRUE,FALSE)`

and so on.

Typically the table `x`

contains several columns of
function values corresponding to different edge corrections.
Auxiliary information for the table identifies one of these
columns as the *recommended value*.
By default, the values of the function `f <- as.function(x)`

are taken from this column of recommended values.
This default can be changed using the argument `value`

,
which can be a character string or character vector of names of
columns of `x`

. Alternatively `value`

can be one of
the abbreviations used by `fvnames`

.

If `value`

specifies a single column of the table,
then the result is a function `f(r)`

with a single numeric
argument `r`

(with the same name as the orginal argument
of the function table).

If `value`

specifies several columns of the table,
then the result is a function `f(r,what)`

where `r`

is the numeric argument and
`what`

is a character string identifying the column of values
to be used.

The formal arguments of the resulting function
are `f(r, what=value)`

, which
means that in a call to this function `f`

, the permissible values
of `what`

are the entries of the original vector `value`

;
the default value of `what`

is the first entry of `value`

.

The command `as.function.fv`

is a method for the generic command
`as.function`

.

`as.function.rhohat`

,
`fv`

,
`fv.object`

,
`fvnames`

,
`plot.fv`

,
`Kest`

```
K <- Kest(cells)
f <- as.function(K)
f
f(0.1)
g <- as.function(K, value=c("iso", "trans"))
g
g(0.1, "trans")
```

Run the code above in your browser using DataLab