Convert to a floatingpoint interval
This functionality does not run in MATLAB.
l
...r
hull(object
)
hull
(object
) returns
a floatingpoint interval enclosing object
.
l ... r
is equivalent to hull(l,
r)
.
hull
converts numerical and interval expressions
to numerical intervals of type DOM_INTERVAL
. It accepts
lists and sets of numerical expressions or intervals as well as numerical
expressions, intervals, and settheoretic functions of intervals and
sets.
Infinities are displayed using RD_INF
for infinity
and RD_NINF
for 
infinity
.
hull
is mapped recursively to the operands
of any expression given—but for subexpressions, lists and sets
are not accepted. Identifiers are replaced by intervals, respecting
a certain subset of properties. Cf. Example 3. Likewise, function
calls and domain elements not overloading hull
are
converted to the interval representing the complex plane.
The output of floatingpoint intervals is influenced by the same parameters as the output of floatingpoint numbers:
The function is sensitive to the environment variable DIGITS
which
determines the numerical working precision.
Each subobject of object
can be evaluated
multiple times and must not have any sideeffects.
hull
returns an interval enclosing its arguments.
You can also use the operator ...
instead of the
function call:
hull(0, PI) = 0 ... PI
Infinities are displayed using RD_NINF
for 
∞ and RD_INF
for infinity:
hull(infinity, 9/7), hull({1/4, 9/7, infinity})
Please note that any number whose absolute value is larger than MuPAD^{®} can store in a float is considered infinite:
hull(0, 1e100000000)^4
Inversion of intervals may lead to unions of intervals. If these
are not required, you may use hull
to unify them:
1/(1 ... 1); hull(%)
The application of hull
to an identifier
without a value returns an interval representing the complex plane:
delete x: hull(x)
Certain properties are respected during this conversion:
assume(x > 0): hull(x); delete x:
This way, you can enclose the values of an expression:
hull(sin(abs(x)))
Calls to "unknown" functions are regarded as potentially returning the complex plane:
hull(f(x))

Arbitrary MuPAD objects 
floatingpoint interval, the
empty set, or FAIL
.
object