Re: Pattern matching
- To: mathgroup at smc.vnet.net
- Subject: [mg116244] Re: Pattern matching
- From: Leonid Shifrin <lshifr at gmail.com>
- Date: Tue, 8 Feb 2011 05:05:56 -0500 (EST)
Bobby,
Yes, indeed, you are right. Thanks for the correction!
Regards,
Leonid
On Mon, Feb 7, 2011 at 9:39 PM, DrMajorBob <btreat1 at austin.rr.com> wrote:
> No... the rule is never applied to fact[-1].
>
> If it were, there WOULD be an infinite loop, since fact[-1] -> -1 fact[-2]
> (not -1 times the previous zero).
>
> The previous zero, multiplied by fact[-1] (unevaluated) is zero, and
> there's no rule for zero, so ReplaceRepeated stops.
>
> Bobby
>
>
> On Mon, 07 Feb 2011 05:07:14 -0600, Leonid Shifrin <lshifr at gmail.com>
> wrote:
>
> This is not a bug. The second rule indeed has no chance to execute. The
>> reason that this does not iterate infinitely is that when the first rule
>> applies for n = 0, we get zero as a result. The next time the rule is
>> applied for n = -1, we get zero again, since we multiply -1 and the
>> previous
>> zero. Since the two consecutive results are the same, ReplaceRepeated
>> stops.
>>
>> I used a similar example in my book, where I made the same statement that
>> we should expect infinite iteration, which is apparently incorrect for
>> this
>> particular problem, as you just pointed out. Will add this to a list of
>> things I have to correct.
>>
>> Regards,
>> Leonid
>>
>>
>> On Sun, Feb 6, 2011 at 1:35 PM, StatsMath <stats.math8 at gmail.com> wrote:
>>
>> Have a question regarding the applicaiton of pattern rules:
>>>
>>> fact[4] //. {fact[n_] :> n fact[n-1], fact[0] -> 1}
>>>
>>> This is a bug since fact[0] needs to be defined 1st, but I expected
>>> the above to grind away idefnitely but it returned a a value 0,
>>> instead of an infininte computation.
>>>
>>> Can you help me understand why the above returns 0?
>>>
>>> Thanks!
>>>
>>>
>>>
>
> --
> DrMajorBob at yahoo.com
>