Merging Trunk into iso-experiment

[Martin Rubey]

Dear Ralf,

I hope you like it...

summary of the test suite:
============================

test cases : 10
tests : 138
tests succeeded : 138
tests failed : 0

This tests also isomorphismtypes of composiition.

I can now go on to think about MultiSorted species. If in luck, I can report on
this on monday.

Many many thanks for your excellent implementation of cycleIndexSeries!

I feel very sorry that I will be unable to provide isomorphismtypes for
functorialComposition...

But: never again delete a file and add it under a different name as you did
with test/setpart.as.nw!

use "svn mv" instead!

Martin


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

[Ralf Hemmecke]

Post by Martin Rubey
============================
test cases : 10
tests : 138
tests succeeded : 138
tests failed : 0
This tests also isomorphismtypes of composiition.

Oh, you made that work... That is wonderful. I'll look at it in a
minute, I will just check in this factorization/prime stuff.

Post by Martin Rubey
But: never again delete a file and add it under a different name as you did
with test/setpart.as.nw!
use "svn mv" instead!

That is exactly what I did. Before you complain, read
http://svnbook.red-bean.com/nightly/en/svn.ref.svn.c.move.html

This command is equivalent to an svn copy followed by svn delete.

So blame subversion, next time.

By the way the ChangeLog clearly says:
* combinat.tex.nw, test/par.as.nw:
Renamed test/setpart.as.nw to test/par.as.nw.

Also svn log shows:

r611 (orig r157): hemmecke | 2007-03-01 14:16:14 +0100

Renamed test/setpart.as.nw to test/par.as.nw.

So you cannot say that I have not given you enough information.

Ralf

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

[Martin Rubey]

Oh, you made that work... That is wonderful. I'll look at it in a minute,

Yes, and thanks to Christian and svn it didn't take more than an half hour!
This AldorUnit stuff is incredibly useful! It would certainly have taken a few
days without it to detect the few mistakes and check whether everything seems
to work.

Thank you!

In fact, I have to admit that I didn't check whether it works in Axiom. I'd
like to think about multisorted species first.

Post by Martin Rubey
But: never again delete a file and add it under a different name as you did
with test/setpart.as.nw!
use "svn mv" instead!

That is exactly what I did. Before you complain, read
http://svnbook.red-bean.com/nightly/en/svn.ref.svn.c.move.html
This command is equivalent to an svn copy followed by svn delete.

Oh, they cheated! This possibility didn't occur to me at all. I am very
sorry. (How often did I write this? Maybe I should do a grep once...)

So you cannot say that I have not given you enough information.

No, the fault is quite on svn's side. Again, sorry...

Two questions:

* should I check it in?

* I did not bother to implement the cycle index series for Subset and
Combination. But I think that's not so pressing currently.

HOWEVER: they return wrong results (namely: 0) currently, since writing
"never" makes the test fail! (Instantiating Subset means that
cycleIndexSeries gets evaluated...)


Martin


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

[Ralf Hemmecke]

Post by Martin Rubey
* should I check it in?

What you do in iso-experiment is up to you. An I would prefer if I see
what you did there before it goes into trunk.

Do you still have the "MultiSet" stuff in or does it now work just with
SetSpecies?

Post by Martin Rubey
* I did not bother to implement the cycle index series for Subset and
Combination. But I think that's not so pressing currently.

But that should be easy. Since Subset = SetSpecies*SetSpecies, you just
have to square the CIS from SetSpecies.

So why don't you simply implement in Subset:

macro cisSet == cycleIndexSeries $ SetSpecies;
cycleIndexSeries: % == cisSet * cisSet;

Shouldn't that be the CIS of Subsets?

(Similar equations for the other two generating series.)

Ralf

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

[Martin Rubey]

Post by Ralf Hemmecke
macro cisSet == cycleIndexSeries $ SetSpecies;
cycleIndexSeries: % == cisSet * cisSet;

won't there be a problem with recursion? Ah, no, only Times depends on Subset,
not the CIS.

OK, I'll do that, but I cannot write TestCases right now.

Done.

Martin


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

[Ralf Hemmecke]

You puzzled me.

I've just commited that factorization/prime stuff, but I don't find
anything new in iso-experiment. Should I see something?

Ralf

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV

[Martin Rubey]

Post by Ralf Hemmecke
You puzzled me.
I've just commited that factorization/prime stuff, but I don't find anything
new in iso-experiment. Should I see something?

I just committed.

Yes, I like your implementation of counting the integer partitions.

Martin


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV