Mini-challenge for RPL enthusiasts
by Joe Horn

It's been forever since the last "mini-challenge".  So here's a really
easy one that everybody can try.

Write the smallest RPL program that gives all the *pairs* of factors
of an integer.

Example:
Input: 12
Output: { { 1 12 } { 2 6 } { 3 4 } }

Another example:
Input: 25
Output: { { 1 25 } { 5 5 } }

Full description: Write an RPL program that inputs any integer and
outputs every possible *pair* of integer factors.  The output is a
list of lists, with each sublist containing a pair of factors, as in
the example above.

Output must be complete.  The order of the factors in each pair, and
the order of the pairs, does not matter.  Repeats are not allowed.

The usual mini-challenge rules apply. Only native User-RPL is
allowed.  Embedding code into a string then executing it via OBJ-> or
STR-> is a Bad Thing.  And so on.

The SMALLEST program (in bytes) wins.  Prize: eternal fame in the
comp.sys.hp48 archives.

-Joe-
http://holyjoe.net/hhc2009

In article 
<d9455264-975c-4243-82ce-60bc7f9248f9@v15g2000prn.googlegroups.com>,
 Joe Horn <joehorn@holyjoe.net> wrote:

> Mini-challenge for RPL enthusiasts
> by Joe Horn
> 
> It's been forever since the last "mini-challenge".  So here's a really
> easy one that everybody can try.
> 
> Write the smallest RPL program that gives all the *pairs* of factors
> of an integer.
> 
> Example:
> Input: 12
> Output: { { 1 12 } { 2 6 } { 3 4 } }
> 
> Another example:
> Input: 25
> Output: { { 1 25 } { 5 5 } }
> 
> Full description: Write an RPL program that inputs any integer and
> outputs every possible *pair* of integer factors.  The output is a
> list of lists, with each sublist containing a pair of factors, as in
> the example above.
> 
> Output must be complete.  The order of the factors in each pair, and
> the order of the pairs, does not matter.  Repeats are not allowed.
> 
> The usual mini-challenge rules apply. Only native User-RPL is
> allowed.  Embedding code into a string then executing it via OBJ-> or
> STR-> is a Bad Thing.  And so on.
> 
> The SMALLEST program (in bytes) wins.  Prize: eternal fame in the
> comp.sys.hp48 archives.
> 
> -Joe-
> http://holyjoe.net/hhc2009

As a first effort:

\<< DIVIS DUP REVLIST 2 \->LIST AXL TRN AXL 1 OVER SIZE 2 / CEIL SUB \>>
# 688Fh
59.5 bytes.

On Jul 30, 12:25=A0am, Virgil <Vir...@home.esc> wrote:
> In article
> <d9455264-975c-4243-82ce-60bc7f924...@v15g2000prn.googlegroups.com>,
> =A0Joe Horn <joeh...@holyjoe.net> wrote:
>
>
>
>
>
> > Mini-challenge for RPL enthusiasts
> > by Joe Horn
>
> > It's been forever since the last "mini-challenge". =A0So here's a reall=
y
> > easy one that everybody can try.
>
> > Write the smallest RPL program that gives all the *pairs* of factors
> > of an integer.
>
> > Example:
> > Input: 12
> > Output: { { 1 12 } { 2 6 } { 3 4 } }
>
> > Another example:
> > Input: 25
> > Output: { { 1 25 } { 5 5 } }
>
> > Full description: Write an RPL program that inputs any integer and
> > outputs every possible *pair* of integer factors. =A0The output is a
> > list of lists, with each sublist containing a pair of factors, as in
> > the example above.
>
> > Output must be complete. =A0The order of the factors in each pair, and
> > the order of the pairs, does not matter. =A0Repeats are not allowed.
>
> > The usual mini-challenge rules apply. Only native User-RPL is
> > allowed. =A0Embedding code into a string then executing it via OBJ-> or
> > STR-> is a Bad Thing. =A0And so on.
>
> > The SMALLEST program (in bytes) wins. =A0Prize: eternal fame in the
> > comp.sys.hp48 archives.
>
> > -Joe-
> >http://holyjoe.net/hhc2009
>
> As a first effort:
>
> \<< DIVIS DUP REVLIST 2 \->LIST AXL TRN AXL 1 OVER SIZE 2 / CEIL SUB \>>
> # 688Fh
> 59.5 bytes.

Virgil, great job - I think you won. My program is over 165 bytes. Is
there actually any room left for improvement?

Thanks for showing teaching me a thing or two about RPL.

And to anyone: more mini-challenges please.

Cheers.

In article 
<a8689525-1fe3-4fbc-b8fb-cd05c0a2f0b5@c14g2000yqm.googlegroups.com>,
 greenchile505 <greenchile505@yahoo.com> wrote:

> On Jul 30, 12:25�am, Virgil <Vir...@home.esc> wrote:
> > In article
> > <d9455264-975c-4243-82ce-60bc7f924...@v15g2000prn.googlegroups.com>,
> > �Joe Horn <joeh...@holyjoe.net> wrote:
> >
> >
> >
> >
> >
> > > Mini-challenge for RPL enthusiasts
> > > by Joe Horn
> >
> > > It's been forever since the last "mini-challenge". �So here's a really
> > > easy one that everybody can try.
> >
> > > Write the smallest RPL program that gives all the *pairs* of factors
> > > of an integer.
> >
> > > Example:
> > > Input: 12
> > > Output: { { 1 12 } { 2 6 } { 3 4 } }
> >
> > > Another example:
> > > Input: 25
> > > Output: { { 1 25 } { 5 5 } }
> >
> > > Full description: Write an RPL program that inputs any integer and
> > > outputs every possible *pair* of integer factors. �The output is a
> > > list of lists, with each sublist containing a pair of factors, as in
> > > the example above.
> >
> > > Output must be complete. �The order of the factors in each pair, and
> > > the order of the pairs, does not matter. �Repeats are not allowed.
> >
> > > The usual mini-challenge rules apply. Only native User-RPL is
> > > allowed. �Embedding code into a string then executing it via OBJ-> or
> > > STR-> is a Bad Thing. �And so on.
> >
> > > The SMALLEST program (in bytes) wins. �Prize: eternal fame in the
> > > comp.sys.hp48 archives.
> >
> > > -Joe-
> > >http://holyjoe.net/hhc2009
> >
> > As a first effort:
> >
> > \<< DIVIS DUP REVLIST 2 \->LIST AXL TRN AXL 1 OVER SIZE 2 / CEIL SUB \>>
> > # 688Fh
> > 59.5 bytes.
> 
> Virgil, great job - I think you won. My program is over 165 bytes. Is
> there actually any room left for improvement?

I found a few alternates in the 60 to 80 byte range, but have not been 
able to improve on my first effort. Very unusual to be able to do so 
well on my first try. But as I have been doing some programs that used 
pretty much all the commands that I used in that first effort, they were 
all in my mind when I began to work on it.
> 
> Thanks for showing teaching me a thing or two about RPL.

I still learn from lots rom others, Joe Horn not least among them.
> 
> And to anyone: more mini-challenges please.
> 
> Cheers.

On Jul 29, 11:50=A0pm, Joe Horn <joeh...@holyjoe.net> wrote:
> Mini-challenge for RPL enthusiasts
> by Joe Horn


It's seems a shame that the first mini challenge to come along in such
a long time is over before it's started. The one submitted solution is
so incredibly elegant that's it's improbable (I hate the word
impossible) to beat. Fun challenge though and fantastic solution. I
learn more about my 50g all the time. :D

Virgil wrote:
> In article 
> <d9455264-975c-4243-82ce-60bc7f9248f9@v15g2000prn.googlegroups.com>,
>  Joe Horn <joehorn@holyjoe.net> wrote:
> 
>> Mini-challenge for RPL enthusiasts
>> by Joe Horn
>>
>> It's been forever since the last "mini-challenge".  So here's a really
>> easy one that everybody can try.
>>
>> Write the smallest RPL program that gives all the *pairs* of factors
>> of an integer.
>>
>> Example:
>> Input: 12
>> Output: { { 1 12 } { 2 6 } { 3 4 } }
>>
>> Another example:
>> Input: 25
>> Output: { { 1 25 } { 5 5 } }
>>
>> Full description: Write an RPL program that inputs any integer and
>> outputs every possible *pair* of integer factors.  The output is a
>> list of lists, with each sublist containing a pair of factors, as in
>> the example above.
>>
>> Output must be complete.  The order of the factors in each pair, and
>> the order of the pairs, does not matter.  Repeats are not allowed.
>>
>> The usual mini-challenge rules apply. Only native User-RPL is
>> allowed.  Embedding code into a string then executing it via OBJ-> or
>> STR-> is a Bad Thing.  And so on.
>>
>> The SMALLEST program (in bytes) wins.  Prize: eternal fame in the
>> comp.sys.hp48 archives.
>>
>> -Joe-
>> http://holyjoe.net/hhc2009
> 
> As a first effort:
> 
> \<< DIVIS DUP REVLIST 2 \->LIST AXL TRN AXL 1 OVER SIZE 2 / CEIL SUB \>>
> # 688Fh
> 59.5 bytes.

You can shorten this one by 2.5 bytes if you use a quirk of the SUB 
function.  It will do the CEIL for you (at least here on my Emu48 HP 
50g), so you can take that out, leaving the slightly shorter:

\<< DIVIS DUP REVLIST 2 \->LIST AXL TRN AXL 1 OVER SIZE 2 / SUB \>>
# C3B5h
57.0 bytes

Credit still goes to Virgil on this one ;-)

-Dave

In article <h577ii$6t0$1@misc-cct.server.rpi.edu>,
 David Brigada <brigad@rpi.edu> wrote:

> Virgil wrote:
> > In article 
> > <d9455264-975c-4243-82ce-60bc7f9248f9@v15g2000prn.googlegroups.com>,
> >  Joe Horn <joehorn@holyjoe.net> wrote:
> > 
> >> Mini-challenge for RPL enthusiasts
> >> by Joe Horn
> >>
> >> It's been forever since the last "mini-challenge".  So here's a really
> >> easy one that everybody can try.
> >>
> >> Write the smallest RPL program that gives all the *pairs* of factors
> >> of an integer.
> >>
> >> Example:
> >> Input: 12
> >> Output: { { 1 12 } { 2 6 } { 3 4 } }
> >>
> >> Another example:
> >> Input: 25
> >> Output: { { 1 25 } { 5 5 } }
> >>
> >> Full description: Write an RPL program that inputs any integer and
> >> outputs every possible *pair* of integer factors.  The output is a
> >> list of lists, with each sublist containing a pair of factors, as in
> >> the example above.
> >>
> >> Output must be complete.  The order of the factors in each pair, and
> >> the order of the pairs, does not matter.  Repeats are not allowed.
> >>
> >> The usual mini-challenge rules apply. Only native User-RPL is
> >> allowed.  Embedding code into a string then executing it via OBJ-> or
> >> STR-> is a Bad Thing.  And so on.
> >>
> >> The SMALLEST program (in bytes) wins.  Prize: eternal fame in the
> >> comp.sys.hp48 archives.
> >>
> >> -Joe-
> >> http://holyjoe.net/hhc2009
> > 
> > As a first effort:
> > 
> > \<< DIVIS DUP REVLIST 2 \->LIST AXL TRN AXL 1 OVER SIZE 2 / CEIL SUB \>>
> > # 688Fh
> > 59.5 bytes.
> 
> You can shorten this one by 2.5 bytes if you use a quirk of the SUB 
> function.  It will do the CEIL for you (at least here on my Emu48 HP 
> 50g), so you can take that out, leaving the slightly shorter:
> 
> \<< DIVIS DUP REVLIST 2 \->LIST AXL TRN AXL 1 OVER SIZE 2 / SUB \>>
> # C3B5h
> 57.0 bytes
> 
> Credit still goes to Virgil on this one ;-)
> 
> -Dave

I knew that somebody could improve on it, somehow. 
Thanks for the tip.
Actually I knew that trick way back when I was new with a 48, way before 
the 49's and 50's, but forgot it over time.

Kind of wish I had held off posting for a few days, just to see what 
others could have come up with.

On Jul 30, 8:25=A0am, Virgil <Vir...@home.esc> wrote:
>
> \<< DIVIS DUP REVLIST 2 \->LIST AXL TRN AXL 1 OVER SIZE 2 / CEIL SUB \>>

This is the first custom program I'm entering into my shiny new HP50g,
but I'm getting a small syntax error. I'd be grateful for some
clarification please...

Looks like my mistake is in this section "2 \->LIST". I'm getting
either a syntax error or AXL error, depending on the various
characters I try in order to get that right.

To me (complete novice at this) it looks like " 2 space backslash
hyphen (minus) rightchevron L I S T " etc. I've tried with hyphens and
underscores and the little right pointing arrow on right shift 0 key
but it's still not right. Anyone help please?

Just to describe properly, program is entered and stored fine. If I
put, say, 30 on the stack and run, then I see two horizontal lists
showing my pairs of factors, but then program stops with error
message  !AXL Error: Bad argument type  '\->LIST'

Thanks if you can point out my silly mistake

On Fri, 04 Sep 2009 23:00:49 -0500:

> syntax error at 2 \->LIST

"\->" means the single "right-arrow" character
(right-shift zero on the calculator keyboard)

Quite a few such 8-bit characters are used in UserRPL programming,
and since these characters do not represent the same symbol
in everyone's computer, there is a "backslash" scheme
for representing these characters in "plainer" (7-bit) text,
for postings, for email, or for external "source" archives:

"Ascii \xx symbols"
http://groups.google.com/group/comp.sys.hp48/msg/52e9cc3ee2b369b8

A web page (by Joe Horn)
containing a larger, easy to read chart:
http://holyjoe.net/hp/tiotable.htm

The calculator itself is able to translate these symbols
between 8-bit internal characters and 7-bit "backslash sequences,"
but only while transferring text files from or to external sources,
or using additional calculator programs, such as:

"Ascii Import/Export for SD card and Emulator"
http://groups.google.com/group/comp.sys.hp48/msg/4e7ed90b3cf11c42

[r->] [OFF]

In article 
<3589290f-b7f7-4246-bd5e-5e9728bede03@r39g2000yqm.googlegroups.com>,
 Grufff <evans.mechanical@googlemail.com> wrote:

> On Jul 30, 8:25�am, Virgil <Vir...@home.esc> wrote:
> >
> > \<< DIVIS DUP REVLIST 2 \->LIST AXL TRN AXL 1 OVER SIZE 2 / CEIL SUB \>>
> 
> This is the first custom program I'm entering into my shiny new HP50g,
> but I'm getting a small syntax error. I'd be grateful for some
> clarification please...
> 
> Looks like my mistake is in this section "2 \->LIST". I'm getting
> either a syntax error or AXL error, depending on the various
> characters I try in order to get that right.


The backslash "\" indicates that the following symbol or symbols 
represent a character that is not representable in the standard newsnet 
character set.

When entering the code into your calculator the sequence "\->" 
represents a simple, and single, right-arrow character, and the whole 
command should appear in the calculator program as ->LIST without any 
backslash.

You might note that the same backslash shown in the \<< and \>> at the 
ends of the program also should appear in the calculator without the 
backslashs.

See if this doesn't clear things up.

On Sep 5, 6:42=A0am, Virgil <Vir...@home.esc> wrote:

> The backslash "\" indicates that the following symbol or symbols
> represent a character that is not representable in the standard newsnet
> character set.
>
> When entering the code into your calculator the sequence "\->"
> represents a simple, and single, right-arrow character, and the whole
> command should appear in the calculator program as ->LIST without any
> backslash.

Oohhh, you're a star! Not only did you write it, but now you're doing
the customer support too.

That fixed things up a treat, and it's now working perfectly. Thanks
very much.

In article 
<8c64f6c9-bcfb-48ce-bf88-89ea876536fd@y36g2000yqh.googlegroups.com>,
 Grufff <evans.mechanical@googlemail.com> wrote:

> On Sep 5, 6:42�am, Virgil <Vir...@home.esc> wrote:
> 
> > The backslash "\" indicates that the following symbol or symbols
> > represent a character that is not representable in the standard newsnet
> > character set.
> >
> > When entering the code into your calculator the sequence "\->"
> > represents a simple, and single, right-arrow character, and the whole
> > command should appear in the calculator program as ->LIST without any
> > backslash.
> 
> Oohhh, you're a star! Not only did you write it, but now you're doing
> the customer support too.
> 
> That fixed things up a treat, and it's now working perfectly. Thanks
> very much.

You are most welcome.

On Sep 5, 6:37=A0am, "John H Meyers" <jhmey...@nomail.invalid> wrote:
> On Fri, 04 Sep 2009 23:00:49 -0500:


> "\->" means the single "right-arrow" character
> (right-shift zero on the calculator keyboard)

> "Ascii \xx symbols"http://groups.google.com/group/comp.sys.hp48/msg/52e9c=
c3ee2b369b8
>
> A web page (by Joe Horn)
> containing a larger, easy to read chart:http://holyjoe.net/hp/tiotable.ht=
m

That's very helpful info, and no doubt will be invaluable as I try and
get to grips with the 50g. Thanks very much.

On Aug 3, 8:42=A0pm, David Brigada <bri...@rpi.edu> wrote:
> Virgil wrote:
> > As a first effort:
>
> > \<< DIVIS DUP REVLIST 2 \->LIST AXL TRN AXL 1 OVER SIZE 2 / CEIL SUB \>=
>
> > # 688Fh
> > 59.5 bytes.
>
> You can shorten this one by 2.5 bytes if you use a quirk of the SUB
> function. =A0It will do the CEIL for you (at least here on my Emu48 HP
> 50g), so you can take that out, leaving the slightly shorter:
>
> \<< DIVIS DUP REVLIST 2 \->LIST AXL TRN AXL 1 OVER SIZE 2 / SUB \>>
> # C3B5h
> 57.0 bytes
>
> Credit still goes to Virgil on this one ;-)
>
> -Dave

My original one was 77 bytes, so hats off to Virgil.  However, besides
not using CEIL, mine also replaced "AXL TRN AXL" with "TRAN".  TRAN
can work directly with a list, so no need for the AXL's.

\<< DIVIS DUP REVLIST 2 \->LIST TRAN 1 OVER SIZE 2 / SUB \>>
# 5B30h
49.0 bytes

-wes

In article 
<cf34db2b-95ea-44b0-88a8-f821d3707f6f@x38g2000yqb.googlegroups.com>,
 Wes <wjltemp-gg@yahoo.com> wrote:

> On Aug 3, 8:42�pm, David Brigada <bri...@rpi.edu> wrote:
> > Virgil wrote:
> > > As a first effort:
> >
> > > \<< DIVIS DUP REVLIST 2 \->LIST AXL TRN AXL 1 OVER SIZE 2 / CEIL SUB \>>
> > > # 688Fh
> > > 59.5 bytes.
> >
> > You can shorten this one by 2.5 bytes if you use a quirk of the SUB
> > function. �It will do the CEIL for you (at least here on my Emu48 HP
> > 50g), so you can take that out, leaving the slightly shorter:
> >
> > \<< DIVIS DUP REVLIST 2 \->LIST AXL TRN AXL 1 OVER SIZE 2 / SUB \>>
> > # C3B5h
> > 57.0 bytes

CEIL and some other commands and functions needing integer arguments 
round such half integers up even on the earliest hp48s.
> >
> > Credit still goes to Virgil on this one ;-)
> >
> > -Dave
> 
> My original one was 77 bytes, so hats off to Virgil.  However, besides
> not using CEIL, mine also replaced "AXL TRN AXL" with "TRAN".  TRAN
> can work directly with a list, so no need for the AXL's.
> 
So it does. But how did you discover it?

> \<< DIVIS DUP REVLIST 2 \->LIST TRAN 1 OVER SIZE 2 / SUB \>>
> # 5B30h
> 49.0 bytes


> 
> -wes

On Sep 5, 7:37=A0pm, A Nony Mouse <T...@mouse.hole> wrote:

> > My original one was 77 bytes, so hats off to Virgil. =A0However, beside=
s
> > not using CEIL, mine also replaced "AXL TRN AXL" with "TRAN". =A0TRAN
> > can work directly with a list, so no need for the AXL's.
>
> So it does. But how did you discover it?

Well, I had "AXL TRAN AXL" in my program and was looking to squeeze a
few more bytes out it.  On a whim I decided to see if TRAN would work
on a list, and voila, it worked.  Afterwords, I looked up the TRAN
command in all the documentation that I have and could not find a
single reference to using TRAN on a list.  Perhaps this is the first!

Nosy shows that TRAN simply converts a list to/from a matrix.

(~xTRAN ROMPTR 314 45)
::
  CK1&Dispatch
  BINT4
  FPTR 6 33B=3D^MATTRAN
  BINT5
  ::
    FPTR 6 17A=3D^LIST2MATRIX
    FPTR 6 33B=3D^MATTRAN
    FPTR 6 179=3D^MATRIX2LIST
  ;
;

-wes

On Sep 5, 8:12=A0pm, Wes <wjltemp...@yahoo.com> wrote:

> Well, I had "AXL TRAN AXL" in my program and was looking to squeeze a
> few more bytes out it. =A0On a whim I decided to see if TRAN would work
> on a list, and voila, it worked. =A0Afterwords, I looked up the TRAN
> command in all the documentation that I have and could not find a
> single reference to using TRAN on a list. =A0Perhaps this is the first!

Make that the second reference.

A little googling reveals that on Feb 23, 2003, John Meyers wrote:
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D
Wow, another slight correction:

TRN (an old HP48 command) does not accept
a "list of lists" format for a matrix,
but TRAN does!
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D
http://groups.google.com/group/comp.sys.hp48/msg/d8b697485305fc32?hl=3Den

-wes

Wes wrote:

> \<< DIVIS DUP REVLIST 2 \->LIST TRAN 1 OVER SIZE 2 / SUB \>>
> # 5B30h
> 49.0 bytes

Dude, you *rock*!  <bows deeply>

-Joe-
27 days until HHC 2009!
http://hhuc.us

On Sep 6, 10:21=A0am, joehorn <joeh...@holyjoe.net> wrote:
> Wes wrote:
> > \<< DIVIS DUP REVLIST 2 \->LIST TRAN 1 OVER SIZE 2 / SUB \>>
> > # 5B30h
> > 49.0 bytes
>
> Dude, you *rock*! =A0<bows deeply>
>
> -Joe-

Bowing to Dave that is.

My own program, weighing in at 77bytes, split up the list into 2 sub-
lists, reversed one of them, then stitched them back together.  Dave's
program more efficiently reverses a copy of the whole list, stitches
these together, then SUBs them.  Using one SUB instead of two saves
quite a bit.

I can't even take credit for the "TRAN with a list" idea, having been
scooped by John by 9 years. :-)

-wes