MathGroup Archive 1993

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

Search the Archive

Re: Thread

  • To: mathgroup at yoda.physics.unc.edu
  • Subject: Re: Thread
  • From: withoff
  • Date: Tue, 23 Nov 93 12:19:18 CST

> Does anyone know why does the Thread function doesn't
> allow a composed header. For example, the following
> works just fine:
> 
>   In[1]:= Thread[f[g[a,b],g[c,d]],g]
> 
>   Out[1]= g[f[a, c], f[b, d]]
> 
>   In[2]:= Head[g[1][a,b]]
> 
>   Out[2]= g[1]
> 
>   In[3]:= Head[g[1][c,d]]
> 
>   Out[3]= g[1]
> 
> But now g[1] is not recognized as an header:
> 
>   In[4]:= Thread[f[g[1][a,b],g[1][c,d]],g[1]]
> 
>   Out[4]= f[g[1][a, b], g[1][c, d]]
> 
> 
>                 --Luis Nunes

This is a previously unreported bug.  Thread thinks the g[1] in
f[g[1][a,b],g[1][c,d]] is not the same as the g[1] in the second
argument of Thread.

If you persuade Mathematica (using Share) to use the same
memory location for both expressions, everything will work
as intended.

In[11]:= t = {f[g[1][a,b],g[1][c,d]],g[1]}

Out[11]= {f[g[1][a, b], g[1][c, d]], g[1]}

In[12]:= Thread @@ t

Out[12]= f[g[1][a, b], g[1][c, d]]

In[13]:= Thread @@ Share[t]

Out[13]= g[1][f[a, c], f[b, d]]

Dave Withoff
Research and Development
Wolfram Research





  • Prev by Date: Missing underscores
  • Next by Date: Fermi integrals
  • Previous by thread: Thread
  • Next by thread: Erroneous factor Pi/4 in Integrate[ 2F1[ 1/2, 3/2, 2, x^2 ], x ]