Re: Comments on Motion 27-A "Decorated Intervals"
Dear Vincent,
thank you for your comments.
Am 15.07.2011 12:14, schrieb Vincent Lefevre:
My comments on Motion 27-A "Decorated Intervals":
Remark 1 "The empty set may get all decorations." is not clear at this
moment. I suppose that the cause is the tracking of the worst possible
decoration (Definition 8).
Yes
Concerning Remark 4 "Applying an arithmetic function to an empty
argument produces an empty set: f(∅) = ∅", you can add "even if
f is a constant function".
Yes
Definitions 7 and 8: the notation for d_f may not be well chosen,
as d_f does not depend only on f, but also on t_1, ..., t_k.
In definition 7 it is stated that d_f is the decoration of the function
f over the bare intervals t_1 ... t_k.
Definition 8: do we really want to always track the worst possible
decoration? Here, if One is the constant function One(x) = 1, then
One(floor([1,3])) would have decoration def, not saf.
Yes, to have a simple rule (KISS)
Section 1.3.2, "1/[ε²,1] = ([1,∞], saf)" is incorrect at Level 1.
That is correct, but should nevertheless be part of the motion...
How is the inclusion defined for the elements of DIR?
In section 1.3.4 we stated that the decoration part is ignored.
According to the examples, the decoration part is ignored (contrary
to Motion 26, which was defining decorations as sets).
About Remark 5 (and the whole motion), what do the decorations really
mean in practice? For instance, consider f(x) = tan(floor(x)*pi+pi/2),
implemented as the composition of the various basic functions, and its
interval extension. At Level 1, it is ndf for X = [1,1.5], but not at
Level 2, where it is con (because of rounding). With X = [1,3], it is
con at both Level 1 and Level 2, though f(x) is nowhere defined. So,
what confidence con brings here?
The decoration con tells you that that the level 2 computation of the
function f(x) may have a singularity on the interval x.
Best regards
Jürgen and Marco
--
o Marco Nehmeier, Lehrstuhl fuer Informatik II
/ \ Universitaet Wuerzburg, Am Hubland, D-97074 Wuerzburg
InfoII o Tel.: +49 931 / 31 88684
/ \ Uni E-Mail: nehmeier@xxxxxxxxxxxxxxxxxxxxxxxxxxx
o o Wuerzburg