MathGroup Archive 2013

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

Search the Archive

Re: Exercise of Programming with Mathematica

  • To: mathgroup at
  • Subject: [mg131857] Re: Exercise of Programming with Mathematica
  • From: Sseziwa Mukasa <mukasa at>
  • Date: Sat, 19 Oct 2013 04:09:56 -0400 (EDT)
  • Delivered-to:
  • Delivered-to:
  • Delivered-to:
  • Delivered-to:
  • References: <>

I can't determine a pattern that represents a pair which does not also match {{a,b},{c,d}}, but I can think of two options.  The first is not to use pattern matching at all, which may be more efficient anyway:

In[5]:= {{a, b}, {c, d}, {e, f}}[[All, {2, 1}]]
{{a, b}, {c, d}}[[All, {2, 1}]]
Out[5]= {{b, a}, {d, c}, {f, e}}
Out[6]= {{b, a}, {d, c}}

The other is to limit the pattern matching to the appropriate level:

In[21]:= Replace[{{a, b}, {c, d}, {e, f}}, {x_, y_} -> {y, x}, {1}]
Replace[{{a, b}, {c, d}}, {x_, y_} -> {y, x}, {1}]
Out[21]= {{b, a}, {d, c}, {f, e}}
Out[22]= {{b, a}, {d, c}}

Elegance as always is in the eye of the beholder,


On Oct 18, 2013, at 4:45 AM, Zhenyi Zhang <infozyzhang at> wrote:

> Here is a rule designed to switch the order of each pair of expressions in a list. It works fine on the
> first example, but fails on the second.
> In[1]:= {{a, b}, {c, d}, {e, f}}/.{x_, y_} :> {y, x}
> Out[1]= {{b, a}, {d, c}, {f, e}}
> In[2]:= {{a, b}, {c, d}}/.{x_, y_} :> {y, x}
> Out[2]= {{c, d}, {a, b}}
> Explain what has gone wrong and rewrite this rule to correct the situation, that is, so that the second
> example returns {{b, a}, {d, c}}
> My solution is the most stupid one.
> {{a, b}, {c, d}}/.{{x_, y_}, {w_, t_}} :> {{y, x}, {t, w}}
> May I ask any elegant solutions?
> Thanks.

  • Prev by Date: Re: Version 9 Trial contaminates Version 8 running paid version
  • Next by Date: Re: Exercise of Programming with Mathematica
  • Previous by thread: Exercise of Programming with Mathematica
  • Next by thread: Re: Exercise of Programming with Mathematica