Maxima: lambdas and sublist_indices behaving strangely when acting on symbols

Thanks very much for the explanation and elegant solution.? I seem to be using an older version of Maxima where gensym is only available as a Lisp function -- after a quick search, I found that using ?gensym() instead of gensym() works on my system.

~ Andrew


Date: Mon, 14 May 2012 15:58:30 +0200
Subject: Re: [Maxima] Maxima: lambdas and sublist_indices behaving strangely when acting on symbols
From: volkervannek at googlemail.com
To: buggi22 at hotmail.com
CC: Maxima at math.utexas.edu

The answer is: lambda quotes its arguments. 

Type 
? lambda
and you get the documentation and examples about lambda.

I believe the following does what you want.

(%i1) f(l, x) ::= buildq([l : l, x : x, z : gensym()], 

sublist_indices(l, lambda([z], z =? x)))$

(%i2) macroexpand(f([x,y,z], a));
(%o2)????? sublist_indices([x, y, z], lambda([g33291], g33291 = a))
(%i3) f([x,y,z], a);
(%o3)???????????????????????????????? []

(%i4) f([x,y,z], x);
(%o4)???????????????????????????????? [1]

HTH
Volker van Nek

2012/5/14 Andy L <buggi22 at hotmail.com>



Hello,

I'm having trouble using sublist_indices on lists of symbols. ?Basically, I'm trying to implement a function that will find the indices at which a given symbol appears within a list of symbols. ?In other words, I want a function f(l,x) which behaves as follows:


? f([x,y,z], x) -> [1]? f([x,y,z], y) -> [2]? f([x,y,z], z) -> [3]? f([x,y,z], a) -> []

Can anyone explain why the following code should yield [1] instead of []?

? f: lambda([l,x], sublist_indices(l, lambda([z], z = x)))$? f([x,y,z], a); ?/* this yields [1] */

The same function behaves as expected when the arguments are strings:

? f(["x","y","z"], "a"); ?/* this yields [], as expected */

An alternate approach reveals some strange behavior:

? k(l,x) := block(? ? [lstrs: map(string, l), xstr: string(x)],? ? display(l,x,lstrs,xstr),? ? sublist_indices(lstrs, lambda([z], z = xstr))? )$? k([x,y,z], a);

Here, l is displayed as "[x,y,z]", but lstrs is displayed as "[a,y,z]". ?I don't understand why "x" gets changed to "a" here, or how to prevent that substitution. ?(True, I could get around this by just changing the parameter "x" to some other unlikely name, like "uncommonDummyVar", but that seems fragile & inelegant.)


Does anyone have any suggestions on how to implement the behavior I'm looking for?



Thanks,

~ Andrew

_______________________________________________

Maxima mailing list

Maxima at math.utexas.edu

http://www.math.utexas.edu/mailman/listinfo/maxima