Re: Re: Bug with Sequence

*To*: mathgroup at smc.vnet.net*Subject*: [mg104656] Re: [mg104633] Re: Bug with Sequence*From*: DrMajorBob <btreat1 at austin.rr.com>*Date*: Fri, 6 Nov 2009 05:15:12 -0500 (EST)*References*: <hcr7ac$8di$1@smc.vnet.net> <200911050851.DAA21614@smc.vnet.net>*Reply-to*: drmajorbob at yahoo.com

On the other hand, "forcing the evaluation" of x below does nothing: x = Sequence[] x = x Sequence[] Sequence[] and yet: x = Sequence[] x = {x} Sequence[] {} Bobby On Thu, 05 Nov 2009 02:51:19 -0600, Szabolcs Horvát <szhorvat at gmail.com> wrote: > On 2009.11.04. 7:34, dh wrote: >> Hello, >> >> has anybody an explanation for the behavior of "Sequence"? I think it is >> >> an ugly bug. >> >> Consider the following that shoud succesively shorten the list t: >> >> >> >> t = {1, 2, 3}; j = 0; >> >> While[ ++j< 4, t[[1]] = Sequence[]; Print[t]] >> >> >> >> this returns: {2,3} three times.Dropping of the first element only seems >> >> to work once. >> >> If you say Information[t] you get: >> >> t={Sequence[],2,3} >> > > Hi Daniel, > > What seems to be happening when > > t[[1]] = x > > is evaluated is that first x is evaluated, then the first element of t > is changed to (the result of the evaluation of) x. But t itself does > not go through an evaluation process. Thus an assignment like t[[1]] is > only capable of changing the first element of t but nothing else. > > If we force t to be evaluated in its entirety after every assignment, > then the program will work as expected: > > While[ ++j < 4, t[[1]] = Sequence[]; t = t; Print[t]] > > The fact that t[[1]] = x type assignments are not able to affect > anything else than the first element is not necessarily a bad thing. > Consider the following (admittedly unusual) example: > > u = Unevaluated[{RandomInteger[], RandomInteger[]}] > > If u[[1]] = 2 would trigger an evaluation of the entire contents of u, > the second element would change too. However, I wouldn't expect that to > happen in this particular example. > > I hope this helps, > Szabolcs > -- DrMajorBob at yahoo.com

**References**:**Re: Bug with Sequence***From:*Szabolcs Horvát <szhorvat@gmail.com>