Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.soft-sys.math.mathematica > #2016

Re: thoughts on how to explain this functionality?

Show all headers | View raw

On Apr 30, 5:52 am, "Scot T. Martin" <smar...@seas.harvard.edu> wrote:
> In[1]:= x = 5
>
> Out[1]= 5
>
> In[2]:= With[{valueQ = If[ValueQ[ReleaseHold[Hold[#]]], #, False] &}, valueQ@x]
>
> Out[2]= 5
>
> In[3]:= With[{valueQ = If[ValueQ[#], #, False] &}, valueQ@x]
>
> Out[3]= False
>
> Anyone have thoughts on why Out[2] and Out[3] are different?
>
> It seems to me that a ReleaseHold[Hold[...]] formulation would net out to no effect, yet the effect is apparent.
>
> My desired output is Out[2] but the formulation of ReleaseHold[Hold[...]] is not elegant.

Scot,

Ive never looked this closely at ValueQ, but it seems to have an
interesting way of evaluating (it effectively checks !
Hold[Evaluate[#]] === Hold[#]). One thing to note is that ValueQ has
the HoldAll attribute, so Hold[ValueQ[#]] actually changes nothing,
then ReleaseHold actually is what changes it. The documentation for
ValueQ says this and then shows a few lines which I think are relate
to your problem. I've copied them below.
ValueQ is HoldAll:

In[1]:= x=y;
In[2]:= {ValueQ[x],ValueQ[y]}
Out[2]= {True,False}

      so above, x has a value, which is y. but y has no value.

Here x is evaluated before ValueQ sees it:
In[3]:= ValueQ /@ {x, y}
Out[3]= {False, False}

     this is the problem you are having, it is not the symbol x which
is passed to your function, but the evaluated version of x (=5), so
you get False

Use Unevaluated to preserve the HoldAll attribute:
In[4]:= ValueQ /@ Unevaluated[{x, y}]
Out[4]= {True, False}

  Hope this helps!
-Stefan Salanski

Back to comp.soft-sys.math.mathematica | Previous | Next — Previous in thread | Find similar | Unroll thread

Thread

thoughts on how to explain this functionality? "Scot T. Martin" <smartin@seas.harvard.edu> - 2011-04-30 09:52 +0000
  Re: thoughts on how to explain this functionality? Stefan <wutchamacallit27@gmail.com> - 2011-05-01 10:21 +0000

csiph-web