Services & Resources / Wolfram Forums / MathGroup Archive
-----

MathGroup Archive 2009

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: Re: part assigned sequence behavior puzzling

  • To: mathgroup at smc.vnet.net
  • Subject: [mg105380] Re: [mg105302] Re: part assigned sequence behavior puzzling
  • From: danl at wolfram.com
  • Date: Mon, 30 Nov 2009 06:11:30 -0500 (EST)
  • References: <hekujm$96g$1@smc.vnet.net>

> I'd say that this is not a wise design decision. IMHO, a statement
> like tmp=tmp should have no effect on tmp, whereas with this
> SequenceHold situation it has.
>
> Cheers -- Sjoerd

Perhaps so. I view it, paraphrasing Churchill, as the worst possible
design, except for all the rest.

In the alternative scenario, Sequence[...] things would (I guess) cause
immediate evaluation of the left hand side. This has some bad consequences
and maybe also raises a few questions.

Bad consequences:

(1) It changes assignment semantics.
(2) It means that deeply nested expressions cannot readily be built
because complexity of adding a level becomes worse (maybe this can largely
be short-circuited, I'm not sure).

Questions:

(1) What is the expected/intended behavior of

In[87]:= ss = {a, b, c, d}; Do[ss[[j]] = Sequence[j, j + 1], {j, Length[ss]}]
?

(2) What about
ss2 = Table[Unevaluated[Sequence[i, i^2]], {i, 4}]
?

Maybe that second one is obvious, and no different from current behavior.

All the same, I think it would not be good to have a semantics wherein the
assignment in the Do loop causes evaluation of the lhs in each iteration.

Daniel Lichtblau
Wolfram Research


> On Nov 26, 6:01 am, Bob Hanlon <hanl... at cox.net> wrote:
>> As stated in the documentation for sequence: assignment and replacement
>> f=
> unctions have attribute SequenceHold
>>
>> Attributes[Set]
>>
>> {HoldFirst,Protected,SequenceHold}
>>
>> tmp = Range[15]
>> tmp[[7]] = Sequence @@ Range[2];
>> tmp[[7]] = Sequence @@ Range[2];
>> tmp[[7]] = Sequence @@ Range[2];
>> tmp
>>
>> {1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}
>>
>> {1,2,3,4,5,6,1,2,8,9,10,11,12,13,14,15}
>>
>> tmp = Range[15]
>> tmp[[7]] = Sequence @@ Range[2];
>> tmp = tmp;
>> tmp[[7]] = Sequence @@ Range[2];
>> tmp = tmp;
>> tmp[[7]] = Sequence @@ Range[2];
>> tmp
>>
>> {1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}
>>
>> {1,2,3,4,5,6,1,2,2,2,8,9,10,11,12,13,14,15}
>>
>> Bob Hanlon
>>
>> ---- mkr <mileskra... at gmail.com> wrote:
>>
>> =============
>> I am puzzled by the following behavior:
>>
>> tmp = Range[15]
>> tmp[[7]] = Sequence @@ Range[2];
>> tmp[[7]] = Sequence @@ Range[2];
>> tmp[[7]] = Sequence @@ Range[2];
>> tmp
>>
>> yields
>>
>> {1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15}
>> {1, 2, 3, 4, 5, 6, 1, 2, 8, 9, 10, 11, 12, 13, 14, 15}
>>
>> I would have expected the repeated assignment to have a repeated
>> effect, thus obtaining
>>
>> {1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15}
>> {1, 2, 3, 4, 5, 6, 1, 2, 2, 2, 8, 9, 10, 11, 12, 13, 14, 15}
>>
>> Where/why am I wrong?
>
>




  • Prev by Date: Re: Re: Bug ??????
  • Next by Date: Re: piecewice pdf, problems with cdf
  • Previous by thread: Re: part assigned sequence behavior puzzling
  • Next by thread: Re: simple question