ojalgo-user

[ojAlgo-user] Markowitz Model Problem

[Ruan W. <is0...@ma...>]

Hi,

I use the  Markowitz model to calculate the optimize some assets' weights,

and set all the assets' lowerlimit with Zero and set all the upperlimits with 

One, but it turns out with some negative value which are little than Zero, it 

seems that the lowerlimit does not work.

So I want to ask what should I do it to confirm the lowerlimit?

--
Best regards

weichao

Re: [ojAlgo-user] Markowitz Model Problem

[Anders P. <an...@op...>]

How small are the negative values?

Give me a unit test (something that I can easily integrate with the  
existing junit tests) that shows what you do, and where/how you think  
it goes wrong.

/Anders


On 14 jan 2009, at 03.58, Ruan Weichao wrote:

>
> Hi,
>
> I use the  Markowitz model to calculate the optimize some assets'  
> weights,
>
> and set all the assets' lowerlimit with Zero and set all the  
> upperlimits with
>
> One, but it turns out with some negative value which are little  
> than Zero, it
>
> seems that the lowerlimit does not work.
>
> So I want to ask what should I do it to confirm the lowerlimit?
>
> --
> Best regards
>
> weichao
>
> ---------------------------------------------------------------------- 
> --------
> This SF.net email is sponsored by:
> SourcForge Community
> SourceForge wants to tell your story.
> http://p.sf.net/sfu/sf-spreadtheword
> _______________________________________________
> ojAlgo-user mailing list
> ojA...@li...
> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
>
>

Re: [ojAlgo-user] Markowitz Model Problem

[<bru...@uk...>]

Hi, even if I am not the expert ni this topic, I think your problem is 
that actually your model fails to find a solution and then just gives you 
wrong results that are beyond your constraints.
I had this problem to when ysing a sytem with ActiveSetSolver.
Before looking at your solution you should check that the optimization 
actually did not failed.

BasicMatrix[] system = new BasicMatrix[6];
//[AE], [be], [Q], [c], [AI], [bi]
.....
//create your matrix here
.....

ActiveSetSolver as = new ActiveSetSolver(system);
 
OptimisationSolver.Result res = as.solve();
State state = res.getState();

if(state != state.FAILED ) { 
        solution = res.getSolution();
}
Let me know if this solve your problem.
Regards
BF





Internet 
is0...@ma...
14/01/2009 02:58
Please respond to
oja...@li...


To
oja...@li...
cc

Subject
[ojAlgo-user] Markowitz Model Problem







Hi,

I use the  Markowitz model to calculate the optimize some assets' weights,

and set all the assets' lowerlimit with Zero and set all the upperlimits 
with 

One, but it turns out with some negative value which are little than Zero, 
it 

seems that the lowerlimit does not work.

So I want to ask what should I do it to confirm the lowerlimit?

--
Best regards

weichao

------------------------------------------------------------------------------
This SF.net email is sponsored by:
SourcForge Community
SourceForge wants to tell your story.
http://p.sf.net/sfu/sf-spreadtheword
_______________________________________________
ojAlgo-user mailing list
ojA...@li...
https://lists.sourceforge.net/lists/listinfo/ojalgo-user



This communication is confidential, may be privileged and is meant only
for the intended recipient. If you are not the intended recipient, please
notify the sender by reply and delete the message from your system.
Any unauthorised dissemination, distribution or copying hereof is
prohibited.

BNP Paribas Trust Corporation UK Limited, BNP Paribas UK Limited, BNP
Paribas Commodity Futures Limited, BNP Paribas Asset Management UK
Limited and Investment Fund Services Limited are authorised and
regulated by the Financial Services Authority.

BNP Paribas London Branch and BNP Paribas Private Bank London
Branch are authorised by the CECEI and supervised by the Commission
Bancaire.

BNP Paribas London Branch is authorised and subject to limited
regulation by the Financial Services Authority.  Details about the extent
of our authorisation and regulation by the Financial Services Authority
are available from us on request. BNP Paribas is also a member of the
London Stock Exchange.

BNP Paribas Private Bank London Branch is subject to limited regulation
by the Financial Services Authority.  Details about the extent of our
authorisation and regulation by the Financial Services Authority are
available from us on request.

BNP Paribas Securities Services London Branch is authorised by the
CECEI and supervised by the AMF, and subject to limited regulation by
the Financial Services Authority.  Details on the extent of our regulation
by the Financial Services Authority are available from us on request.
BNP Paribas Securities Services is also a member of the London Stock
Exchange.

Re: [ojAlgo-user] Markowitz Model Problem

[Anders P. <an...@op...>]

I believe this is the first time a list member (other than me) tried  
to help another member!

On 14 jan 2009, at 09.26, bru...@uk... wrote:

> Hi, even if I am not the expert ni this topic, I think your problem  
> is that actually your model fails to find a solution and then just  
> gives you wrong results that are beyond your constraints.

Please give me test cases for things you think should work, but  
currently doesn't.

> I had this problem to when ysing a sytem with ActiveSetSolver.
> Before looking at your solution you should check that the  
> optimization actually did not failed.

This could be the case, but I'd like to see the actual case. With all  
lower limits set to 0 and all upper limits set to 1 (and no other  
constraints) the ActiveSetSolver should be able to find a solution.

The quadratic solvers has been much improved since the release of  
version 25. If it's not to much trouble I'd recommend getting the  
latest source code from CVS. There's still  a few things I'd like to  
fix before releasing version 26.

/Anders


> BasicMatrix[] system = new BasicMatrix[6];
> //[AE], [be], [Q], [c], [AI], [bi]
> .....
> //create your matrix here
> .....
>
> ActiveSetSolver as = new ActiveSetSolver(system);
>
> OptimisationSolver.Result res = as.solve();
> State state = res.getState();
>
> if(state != state.FAILED ) {
>         solution = res.getSolution();
> }
> Let me know if this solve your problem.
> Regards
> BF
>
>
>
>
> Internet
> is0...@ma...
> 14/01/2009 02:58
> Please respond to
> oja...@li...
>
> To
> oja...@li...
> cc
> Subject
> [ojAlgo-user] Markowitz Model Problem
>
>
>
>
>
>
> Hi,
>
> I use the  Markowitz model to calculate the optimize some assets'  
> weights,
>
> and set all the assets' lowerlimit with Zero and set all the  
> upperlimits with
>
> One, but it turns out with some negative value which are little  
> than Zero, it
>
> seems that the lowerlimit does not work.
>
> So I want to ask what should I do it to confirm the lowerlimit?
>
> --
> Best regards
>
> weichao
>
> ---------------------------------------------------------------------- 
> --------
> This SF.net email is sponsored by:
> SourcForge Community
> SourceForge wants to tell your story.
> http://p.sf.net/sfu/sf-spreadtheword
> _______________________________________________
> ojAlgo-user mailing list
> ojA...@li...
> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
>
> This communication is confidential, may be privileged and is meant  
> only for the intended recipient. If you are not the intended  
> recipient, please notify the sender by reply and delete the message  
> from your system. Any unauthorised dissemination, distribution or  
> copying hereof is prohibited. BNP Paribas Trust Corporation UK  
> Limited, BNP Paribas UK Limited, BNP Paribas Commodity Futures  
> Limited, BNP Paribas Asset Management UK Limited and Investment  
> Fund Services Limited are authorised and regulated by the Financial  
> Services Authority. BNP Paribas London Branch and BNP Paribas  
> Private Bank London Branch are authorised by the CECEI and  
> supervised by the Commission Bancaire. BNP Paribas London Branch is  
> authorised and subject to limited regulation by the Financial  
> Services Authority. Details about the extent of our authorisation  
> and regulation by the Financial Services Authority are available  
> from us on request. BNP Paribas is also a member of the London  
> Stock Exchange. BNP Paribas Private Bank London Branch is subject  
> to limited regulation by the Financial Services Authority. Details  
> about the extent of our authorisation and regulation by the  
> Financial Services Authority are available from us on request. BNP  
> Paribas Securities Services London Branch is authorised by the  
> CECEI and supervised by the AMF, and subject to limited regulation  
> by the Financial Services Authority. Details on the extent of our  
> regulation by the Financial Services Authority are available from  
> us on request. BNP Paribas Securities Services is also a member of  
> the London Stock Exchange.
> ---------------------------------------------------------------------- 
> --------
> This SF.net email is sponsored by:
> SourcForge Community
> SourceForge wants to tell your story.
> http://p.sf.net/sfu/sf- 
> spreadtheword_______________________________________________
> ojAlgo-user mailing list
> ojA...@li...
> https://lists.sourceforge.net/lists/listinfo/ojalgo-user

Re: [ojAlgo-user] Markowitz Model Problem

[Ruan W. <is0...@ma...>]

<HTML>
<HEAD>
<META content="text/html; charset=gb2312" http-equiv=Content-Type>
<META content="OPENWEBMAIL" name=GENERATOR>
</HEAD>
<BODY bgColor=#ffffff>
<font size="2">
<p>
<br />Thank you for your response. 
<br />
<br />Here's my model's detail: 
<br />
<br />[AE]={1.0, &nbsp;&nbsp;&nbsp;&nbsp;1.0, &nbsp;&nbsp;&nbsp;&nbsp;1.0, &nbsp;&nbsp; &nbsp;&nbsp;1.0, &nbsp;&nbsp;&nbsp;&nbsp;1.0, &nbsp;&nbsp;&nbsp;&nbsp;1.0, &nbsp; &nbsp;&nbsp;&nbsp; 1.0} </p>

<p>
<br />[be]={1.0} </p>

<p>
<br />[Q]={3.048907897157133E-4,&nbsp;1.6671472561019247E-4,&nbsp;4.4500080981934345E-4,&nbsp;-5.389129745055723E-4,&nbsp;-2.6090705011393183E-4,&nbsp;-1.2633284900760366E-4,&nbsp;-6.485428846447651E-7},
<br />{1.6671472561019247E-4,&nbsp;2.341985572849691E-4,&nbsp;2.9113916450678265E-4,&nbsp;-4.5760873539850514E-4,&nbsp;1.3078636134987255E-5,&nbsp;-2.354289901013046E-5,&nbsp;-7.578030042426654E-7},
<br />{4.4500080981934345E-4,&nbsp;2.9113916450678265E-4,&nbsp;7.46023915996829E-4,&nbsp;-0.0010247176498305568,&nbsp;-2.6745504327902895E-4,&nbsp;-1.6563544154823496E-4,&nbsp;-8.293698990696063E-7},
<br />{-5.389129745055723E-4,&nbsp;-4.5760873539850514E-4,&nbsp;-0.0010247176498305568,&nbsp;0.001754169535149865,&nbsp;2.0293065310212377E-4,&nbsp;2.1401092557826588E-4,&nbsp;1.0252846778608953E-7},
<br />{-2.6090705011393183E-4,&nbsp;1.3078636134987255E-5,&nbsp;-2.6745504327902895E-4,&nbsp;2.0293065310212377E-4,&nbsp;4.632320892679136E-4,&nbsp;1.7969731066037214E-4,&nbsp;2.4953495129362833E-8},
<br />{-1.2633284900760366E-4,&nbsp;-2.354289901013046E-5,&nbsp;-1.6563544154823496E-4,&nbsp;2.1401092557826588E-4,&nbsp;1.7969731066037214E-4,&nbsp;8.346410612364995E-5,&nbsp;-7.02099350897589E-8},
<br />{-6.485428846447651E-7,&nbsp;-7.578030042426654E-7,&nbsp;-8.293698990696063E-7,&nbsp;1.0252846778608953E-7,&nbsp;2.4953495129362833E-8,&nbsp;-7.02099350897589E-8,&nbsp;8.367244992498656E-9}}</p>

<p>[c]={{0.010638291263564232}, 
<br />{0.013500370827906071}, 
<br />{0.011390037735101773}, 
<br />{0.010385042339767682}, 
<br />{3.812208389845893E-4}, 
<br />{0.002315505853720011}, 
<br />{0.0}} </p>

<p>
<br />[AI]={{1.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp; &nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp; &nbsp;&nbsp; 0.0}, 
<br />{0.0, &nbsp;&nbsp;&nbsp;&nbsp;1.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp; &nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp; &nbsp;&nbsp; 0.0}, 
<br />{0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;1.0, &nbsp;&nbsp;&nbsp; &nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp; &nbsp;&nbsp; 0.0}, 
<br />{0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp; &nbsp;1.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp; &nbsp;&nbsp; 0.0}, 
<br />{0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp; &nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;1.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp; &nbsp;&nbsp; 0.0}, 
<br />{0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp; &nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;1.0, &nbsp;&nbsp; &nbsp;&nbsp; 0.0}, 
<br />{0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp; &nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp; &nbsp;&nbsp; 1.0}, 
<br />{-1.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp; &nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp; &nbsp;&nbsp; 0.0}, 
<br />{0.0, &nbsp;&nbsp;&nbsp;&nbsp;-1.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp; &nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp; &nbsp;&nbsp; 0.0}, 
<br />{0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;-1.0, &nbsp;&nbsp;&nbsp; &nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp; &nbsp;&nbsp; 0.0}, 
<br />{0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp; &nbsp;-1.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp; &nbsp;&nbsp; 0.0}, 
<br />{0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp; &nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;-1.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp; &nbsp;&nbsp; 0.0}, 
<br />{0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp; &nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;-1.0, &nbsp;&nbsp; &nbsp;&nbsp; 0.0}, 
<br />{0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp; &nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp;&nbsp;&nbsp;0.0, &nbsp;&nbsp; &nbsp;&nbsp; -1.0}} 
<br /></p>

<p>[bi]={{1.0}, 
<br />{1.0}, 
<br />{1.0}, 
<br />{1.0}, 
<br />{1.0}, 
<br />{1.0}, 
<br />{1.0}, 
<br />{0.0}, 
<br />{0.0}, 
<br />{0.0}, 
<br />{0.0}, 
<br />{0.0}, 
<br />{0.0}, 
<br />{0.0}} 
<br /></p>

<p>And here's the result:</p>

<p>-0.25
<br />-0.25
<br />-0.25
<br />-0.25
<br />0.75
<br />0.75
<br />0.75</p>

<p>And i found that if i change the nagetive ExceptedReturn to positive, the result will be fine.</p>

<p>
<br />On Wed, 14 Jan 2009 09:27:53 +0100, Anders Peterson wrote 
<br />&gt; How small are the negative values? 
<br />&gt; 
<br />&gt; Give me a unit test (something that I can easily integrate with the &nbsp; 
<br />&gt; existing junit tests) that shows what you do, and where/how you 
<br />&gt; think &nbsp;it goes wrong. 
<br />&gt; 
<br />&gt; /Anders 
<br />&gt; 
<br />&gt; On 14 jan 2009, at 03.58, Ruan Weichao wrote: 
<br />&gt; 
<br />&gt; &gt; 
<br />&gt; &gt; Hi, 
<br />&gt; &gt; 
<br />&gt; &gt; I use the &nbsp;Markowitz model to calculate the optimize some assets' &nbsp; 
<br />&gt; &gt; weights, 
<br />&gt; &gt; 
<br />&gt; &gt; and set all the assets' lowerlimit with Zero and set all the &nbsp; 
<br />&gt; &gt; upperlimits with 
<br />&gt; &gt; 
<br />&gt; &gt; One, but it turns out with some negative value which are little &nbsp; 
<br />&gt; &gt; than Zero, it 
<br />&gt; &gt; 
<br />&gt; &gt; seems that the lowerlimit does not work. 
<br />&gt; &gt; 
<br />&gt; &gt; So I want to ask what should I do it to confirm the lowerlimit? 
<br />&gt; &gt; 
<br />&gt; &gt; -- 
<br />&gt; &gt; Best regards 
<br />&gt; &gt; 
<br />&gt; &gt; weichao 
<br />&gt; &gt; 
<br />&gt; &gt; ---------------------------------------------------------------------- 
<br />&gt; &gt; -------- 
<br />&gt; &gt; This SF.net email is sponsored by: 
<br />&gt; &gt; SourcForge Community 
<br />&gt; &gt; SourceForge wants to tell your story. 
<br />&gt; &gt; <a href="http://p.sf.net/sfu/sf-spreadtheword" target="_blank">http://p.sf.net/sfu/sf-spreadtheword</a> 
<br />&gt; &gt; _______________________________________________ 
<br />&gt; &gt; ojAlgo-user mailing list 
<br />&gt; &gt; ojA...@li... 
<br />&gt; &gt; <a href="https://lists.sourceforge.net/lists/listinfo/ojalgo-user" target="_blank">https://lists.sourceforge.net/lists/listinfo/ojalgo-user</a> 
<br />&gt; &gt; 
<br />&gt; &gt; 
<br />&gt; 
<br />&gt; ---------------------------------------------------------------------------- 
<br />-- 
<br />&gt; This SF.net email is sponsored by: 
<br />&gt; SourcForge Community 
<br />&gt; SourceForge wants to tell your story. 
<br />&gt; <a href="http://p.sf.net/sfu/sf-spreadtheword" target="_blank">http://p.sf.net/sfu/sf-spreadtheword</a> 
<br />&gt; _______________________________________________ 
<br />&gt; ojAlgo-user mailing list 
<br />&gt; ojA...@li... 
<br />&gt; <a href="https://lists.sourceforge.net/lists/listinfo/ojalgo-user" target="_blank">https://lists.sourceforge.net/lists/listinfo/ojalgo-user</a> 
<br />
<br />
<br />-- 
<br />Best regards 
<br /></p>
</font>
</BODY>
</HTML>

Re: [ojAlgo-user] Markowitz Model Problem

[Anders P. <an...@op...>]

You have to send code (snippets) that I can execute.

Are you using v25 or the latest from cvs?

How do you instantiate the ActiveSetSolver?

What's the correct solution?

You say that if you change the sign of "C" you get an ok result. What  
does this mean? Did you make a mistake with the sign, or do you  
suspect there is a bug in ojAlgo?

I need better info regarding what you do, and what the problem is!

/Anders


On 15 jan 2009, at 06.14, Ruan Weichao wrote:

>
> Thank you for your response.
>
> Here's my model's detail:
>
> [AE]={1.0,     1.0,     1.0,      1.0,     1.0,     1.0,       1.0}
>
>
> [be]={1.0}
>
>
> [Q]={3.048907897157133E-4, 1.6671472561019247E-4,  
> 4.4500080981934345E-4, -5.389129745055723E-4,  
> -2.6090705011393183E-4, -1.2633284900760366E-4,  
> -6.485428846447651E-7},
> {1.6671472561019247E-4, 2.341985572849691E-4,  
> 2.9113916450678265E-4, -4.5760873539850514E-4,  
> 1.3078636134987255E-5, -2.354289901013046E-5, -7.578030042426654E-7},
> {4.4500080981934345E-4, 2.9113916450678265E-4, 7.46023915996829E-4,  
> -0.0010247176498305568, -2.6745504327902895E-4,  
> -1.6563544154823496E-4, -8.293698990696063E-7},
> {-5.389129745055723E-4, -4.5760873539850514E-4,  
> -0.0010247176498305568, 0.001754169535149865,  
> 2.0293065310212377E-4, 2.1401092557826588E-4, 1.0252846778608953E-7},
> {-2.6090705011393183E-4, 1.3078636134987255E-5,  
> -2.6745504327902895E-4, 2.0293065310212377E-4,  
> 4.632320892679136E-4, 1.7969731066037214E-4, 2.4953495129362833E-8},
> {-1.2633284900760366E-4, -2.354289901013046E-5,  
> -1.6563544154823496E-4, 2.1401092557826588E-4,  
> 1.7969731066037214E-4, 8..346410612364995E-5, -7.02099350897589E-8},
> {-6.485428846447651E-7, -7.578030042426654E-7,  
> -8.293698990696063E-7, 1.0252846778608953E-7,  
> 2.4953495129362833E-8, -7.02099350897589E-8, 8.367244992498656E-9}}
>
> [c]={{0.010638291263564232},
> {0.013500370827906071},
> {0.011390037735101773},
> {0.010385042339767682},
> {3.812208389845893E-4},
> {0.002315505853720011},
> {0.0}}
>
>
> [AI]={{1.0,     0.0,     0.0,      0.0,     0.0,     0.0,       0.0},
> {0.0,     1.0,     0.0,      0.0,     0.0,     0.0,       0.0},
> {0.0,     0.0,     1.0,      0.0,     0.0,     0.0,       0.0},
> {0.0,     0.0,     0.0,      1.0,     0.0,     0.0,       0.0},
> {0.0,     0.0,     0.0,      0.0,     1.0,     0.0,       0.0},
> {0.0,     0.0,     0.0,      0.0,     0.0,     1.0,       0.0},
> {0.0,     0.0,     0.0,      0.0,     0.0,     0.0,       1.0},
> {-1.0,     0.0,     0.0,      0.0,     0.0,     0.0,       0.0},
> {0.0,     -1.0,     0.0,      0.0,     0.0,     0.0,       0.0},
> {0.0,     0.0,     -1.0,      0.0,     0.0,     0.0,       0.0},
> {0.0,     0.0,     0.0,      -1.0,     0.0,     0.0,       0.0},
> {0.0,     0.0,     0.0,      0.0,     -1.0,     0.0,       0.0},
> {0.0,     0.0,     0.0,      0.0,     0.0,     -1.0,       0.0},
> {0.0,     0.0,     0.0,      0.0,     0.0,     0.0,       -1.0}}
>
> [bi]={{1.0},
> {1.0},
> {1.0},
> {1.0},
> {1.0},
> {1.0},
> {1.0},
> {0.0},
> {0.0},
> {0.0},
> {0.0},
> {0.0},
> {0.0},
> {0.0}}
>
> And here's the result:
>
> -0.25
> -0.25
> -0.25
> -0.25
> 0.75
> 0.75
> 0.75
>
> And i found that if i change the nagetive ExceptedReturn to  
> positive, the result will be fine.
>
>
> On Wed, 14 Jan 2009 09:27:53 +0100, Anders Peterson wrote
> > How small are the negative values?
> >
> > Give me a unit test (something that I can easily integrate with the
> > existing junit tests) that shows what you do, and where/how you
> > think  it goes wrong.
> >
> > /Anders
> >
> > On 14 jan 2009, at 03.58, Ruan Weichao wrote:
> >
> > >
> > > Hi,
> > >
> > > I use the  Markowitz model to calculate the optimize some assets'
> > > weights,
> > >
> > > and set all the assets' lowerlimit with Zero and set all the
> > > upperlimits with
> > >
> > > One, but it turns out with some negative value which are little
> > > than Zero, it
> > >
> > > seems that the lowerlimit does not work.
> > >
> > > So I want to ask what should I do it to confirm the lowerlimit?
> > >
> > > --
> > > Best regards
> > >
> > > weichao
> > >
> > >  
> ----------------------------------------------------------------------
> > > --------
> > > This SF.net email is sponsored by:
> > > SourcForge Community
> > > SourceForge wants to tell your story.
> > > http://p.sf.net/sfu/sf-spreadtheword
> > > _______________________________________________
> > > ojAlgo-user mailing list
> > > ojA...@li...
> > > https://lists.sourceforge.net/lists/listinfo/ojalgo-user
> > >
> > >
> >
> >  
> ---------------------------------------------------------------------- 
> ------
> -- 
> > This SF.net email is sponsored by:
> > SourcForge Community
> > SourceForge wants to tell your story.
> > http://p.sf.net/sfu/sf-spreadtheword
> > _______________________________________________
> > ojAlgo-user mailing list
> > ojA...@li...
> > https://lists.sourceforge.net/lists/listinfo/ojalgo-user
>
>
> -- 
> Best regards
>
> ---------------------------------------------------------------------- 
> --------
> This SF.net email is sponsored by:
> SourcForge Community
> SourceForge wants to tell your story.
> http://p.sf.net/sfu/sf- 
> spreadtheword_______________________________________________
> ojAlgo-user mailing list
> ojA...@li...
> https://lists.sourceforge.net/lists/listinfo/ojalgo-user

Re: [ojAlgo-user] Markowitz Model Problem

[Ruan W. <is0...@ma...>]

<HTML>
<HEAD>
<META content="text/html; charset=gb2312" http-equiv=Content-Type>
<META content="OPENWEBMAIL" name=GENERATOR>
</HEAD>
<BODY bgColor=#ffffff>
<font size="2">
<p><font size="2"><b>Hi，i am using v25 and here's my code:</b></font></p>

<p><strong><font size="2">/********************************************************************************/</font></strong></p>

<p>import java.math.BigDecimal;
<br />import java.util.List;</p>

<p>import org.ojalgo.finance.portfolio.MarketEquilibrium;
<br />import org.ojalgo.finance.portfolio.MarkowitzModel;
<br />import org.ojalgo.matrix.*;
<br />import org.ojalgo.matrix.store.BigDenseStore;
<br />import org.ojalgo.matrix.store.MatrixStore;</p>

<p>
<br />public class TestMarkowitz {
<br />&nbsp;
<br />&nbsp;public double getCovarance(double[] value1, double[] value2){
<br />&nbsp;&nbsp;
<br />&nbsp;&nbsp;int n = value1.length;&nbsp;
<br />&nbsp;&nbsp;double averOne = 0;
<br />&nbsp;&nbsp;double averTwo = 0;
<br />&nbsp;&nbsp;for(int i=0;i&lt;n;i++)
<br />&nbsp;&nbsp;{
<br />&nbsp;&nbsp;&nbsp;averOne+=value1[i];
<br />&nbsp;&nbsp;&nbsp;averTwo+=value2[i];
<br />&nbsp;&nbsp;}
<br />&nbsp;&nbsp;averOne /= n;
<br />&nbsp;&nbsp;averTwo /= n;
<br />&nbsp;&nbsp;
<br />&nbsp;&nbsp;double sum = 0;&nbsp;&nbsp;
<br />&nbsp;&nbsp;for(int i=0;i&lt;n;i++)
<br />&nbsp;&nbsp;&nbsp;sum+=(value1[i]-averOne)*(value2[i]-averTwo);</p>

<p>&nbsp;&nbsp;return sum / (n-1);
<br />&nbsp;}
<br />&nbsp;
<br />&nbsp;public BasicMatrix getACovariances(double[][] returns){
<br />&nbsp;&nbsp;
<br />&nbsp;&nbsp;int row = returns.length;
<br />&nbsp;&nbsp;int col = returns[0].length;
<br />&nbsp;&nbsp;MatrixStore&lt;BigDecimal&gt; a = BigDenseStore.FACTORY.makeZero(row, col);
<br />&nbsp;&nbsp;BasicMatrix covariances = new BigMatrix(a);
<br />&nbsp;&nbsp;
<br />&nbsp;&nbsp;for(int i=1;i&lt;=row;i++){
<br />&nbsp;&nbsp;&nbsp;for(int j=i;j&lt;=col;j++){&nbsp;&nbsp;&nbsp;&nbsp;
<br />&nbsp;&nbsp;&nbsp;&nbsp;double tmp = this.getCovarance(returns[i-1], returns[j-1]);
<br />&nbsp;&nbsp;&nbsp;&nbsp;covariances = covariances.set(i-1, j-1,tmp);
<br />&nbsp;&nbsp;&nbsp;&nbsp;covariances = covariances.set(j-1, i-1,tmp);
<br />&nbsp;&nbsp;&nbsp;}
<br />&nbsp;&nbsp;}
<br />&nbsp;&nbsp;return covariances;
<br />&nbsp;}
<br />&nbsp;
<br />&nbsp;public BasicMatrix getAnExpectedExcessReturns(double[][] returns){
<br />&nbsp;&nbsp;
<br />&nbsp;&nbsp;int row = returns.length;
<br />&nbsp;&nbsp;int col = returns[0].length;
<br />&nbsp;&nbsp;
<br />&nbsp;&nbsp;MatrixStore&lt;BigDecimal&gt; a = BigDenseStore.FACTORY.makeZero(row, 1);
<br />&nbsp;&nbsp;BasicMatrix expectedExcessReturns = new BigMatrix(a);
<br />&nbsp;&nbsp;
<br />&nbsp;&nbsp;double riskFreeReturn = 0;
<br />&nbsp;&nbsp;for(int i=1;i&lt;=col;i++)
<br />&nbsp;&nbsp;&nbsp;riskFreeReturn+=returns[row-1][i-1];
<br />&nbsp;&nbsp;riskFreeReturn/=row;
<br />&nbsp;&nbsp;
<br />&nbsp;&nbsp;for(int i=1;i&lt;=row;i++){
<br />&nbsp;&nbsp;&nbsp;
<br />&nbsp;&nbsp;&nbsp;double tmp = 0;
<br />&nbsp;&nbsp;&nbsp;for(int j=1;j&lt;=col;j++)
<br />&nbsp;&nbsp;&nbsp;&nbsp;tmp+=returns[i-1][j-1];
<br />&nbsp;&nbsp;&nbsp;tmp/=col;
<br />&nbsp;&nbsp;&nbsp;
<br />&nbsp;&nbsp;&nbsp;expectedExcessReturns = expectedExcessReturns.set(i-1, 0, tmp-riskFreeReturn);&nbsp;&nbsp;
<br />&nbsp;&nbsp;}
<br />&nbsp;&nbsp;
<br />&nbsp;&nbsp;return expectedExcessReturns;
<br />&nbsp;}</p>

<p>&nbsp;public static void main(String[] args)
<br />&nbsp;{&nbsp;&nbsp;
<br />&nbsp;&nbsp;int assetNum=7;
<br />&nbsp;&nbsp;double[][] assets_return={
<br />&nbsp;&nbsp;&nbsp;{-1.5905837442343828E-4, -0.03062360801781757, -0.029857534032853142, -0.011811692726036832, -0.017972310602803136, 0.017338003502626997, 0.0},
<br />&nbsp;&nbsp;&nbsp;{-0.02757158006362653, -0.02562704471101405, -0.011751538891997735, -0.024915062287655786, -0.01684088269454123, 0.013585351447135364, 0.0},
<br />&nbsp;&nbsp;&nbsp;{-0.00699300699300693, -0.033802816901408676, -0.04675196850393671, -0.021166752710376546, -0.007911392405063583, 0.03827751196172254, 0.0},
<br />&nbsp;&nbsp;&nbsp;{-0.007626310772164015, 0.0038424591738713027, 0.02488038277511978, 0.025210084033613675, -0.02003642987249557, -0.09758364312267642, 0.0},
<br />&nbsp;&nbsp;&nbsp;{-0.03965053763440893, 0.021693491952414375, 0.01643835616438392, -0.007412398921833087, 0.01765105227427014, -0.010006671114076025, 0.0},
<br />&nbsp;&nbsp;&nbsp;{-0.017821782178217872, 0.005040322580645311, 0.006018054162487363, 9.008107296569024E-4, 0.002999999999999824, -0.01196410767696908, 0.0},
<br />&nbsp;&nbsp;&nbsp;{2.630552127527583E-4, 2.5867028174649627E-4, 2.3866431891514327E-4, 1.9564035993080523E-4, 2.351016690966669E-4, 1.9070675120065465E-4, 0.0}
<br />&nbsp;&nbsp;};
<br />&nbsp;&nbsp;&nbsp;&nbsp;
<br />&nbsp;&nbsp;TestMarkowitz tm = new TestMarkowitz();
<br />&nbsp;&nbsp;BasicMatrix covariances = tm.getACovariances(assets_return);
<br />&nbsp;&nbsp;BasicMatrix expectedExcessReturns = tm.getAnExpectedExcessReturns(assets_return);&nbsp;&nbsp;
<br />&nbsp;&nbsp;BigDecimal riskAversion = new BigDecimal(1.0);
<br />&nbsp;&nbsp;
<br />&nbsp;&nbsp;MarketEquilibrium marketEquilibrium = new MarketEquilibrium(covariances,riskAversion);
<br />&nbsp;&nbsp;MarkowitzModel markowitzModel = new MarkowitzModel(marketEquilibrium,expectedExcessReturns);
<br />&nbsp;&nbsp;
<br />&nbsp;&nbsp;for(int i=0;i&lt;assetNum;i++){
<br />&nbsp;&nbsp;&nbsp;markowitzModel.setLowerLimit(i, new BigDecimal(0.0));
<br />&nbsp;&nbsp;&nbsp;markowitzModel.setUpperLimit(i, new BigDecimal(1.0));
<br />&nbsp;&nbsp;}
<br />&nbsp;&nbsp;List&lt;BigDecimal&gt; re = markowitzModel.getWeights();
<br />&nbsp;&nbsp;
<br />&nbsp;&nbsp;System.out.println(&quot;=======result====================&quot;);
<br />&nbsp;&nbsp;for(int i=0;i&lt;re.size();i++)
<br />&nbsp;&nbsp;&nbsp;System.out.println(re.get(i));
<br />&nbsp;&nbsp;
<br />&nbsp;&nbsp;return;
<br />&nbsp;}
<br />&nbsp;
<br />}
<br /></p>

<p><strong><font size="2">/********************************************************************************/</font></strong></p>

<p><font size="2"><b>and the result turn out as:</b></font></p>

<p>-0.25
<br />-0.25
<br />-0.25
<br />-0.25
<br />0.75
<br />0.75
<br />0.75</p>

<p><strong><font size="2">/****************************************************************************/</font></strong></p>

<p><strong><font size="2">If you find out any problem, just let me know , thanks a lot!</font></strong></p>
</font>
<p><font size="2">On Thu, 15 Jan 2009 09:20:52 +0100, Anders Peterson wrote 
<br />&gt; You have to send code (snippets) that I can execute. 
<br />&gt; 
<br />&gt; Are you using v25 or the latest from cvs? 
<br />&gt; 
<br />&gt; How do you instantiate the ActiveSetSolver? 
<br />&gt; 
<br />&gt; What's the correct solution? 
<br />&gt; 
<br />&gt; You say that if you change the sign of &quot;C&quot; you get an ok result. 
<br />&gt; What &nbsp;does this mean? Did you make a mistake with the sign, or do 
<br />&gt; you &nbsp;suspect there is a bug in ojAlgo? 
<br />&gt; 
<br />&gt; I need better info regarding what you do, and what the problem is! 
<br />&gt; 
<br />&gt; /Anders 
<br />&gt; 
<br />&gt; On 15 jan 2009, at 06.14, Ruan Weichao wrote: 
<br />&gt; 
<br />&gt; &gt; 
<br />&gt; &gt; Thank you for your response. 
<br />&gt; &gt; 
<br />&gt; &gt; Here's my model's detail: 
<br />&gt; &gt; 
<br />&gt; &gt; [AE]={1.0, &nbsp; &nbsp; 1.0, &nbsp; &nbsp; 1.0, &nbsp; &nbsp; &nbsp; 1.0, &nbsp; &nbsp; 1.0, &nbsp; &nbsp; 1.0, &nbsp; &nbsp; &nbsp; 1.0} 
<br />&gt; &gt; 
<br />&gt; &gt; 
<br />&gt; &gt; [be]={1.0} 
<br />&gt; &gt; 
<br />&gt; &gt; 
<br />&gt; &gt; [Q]={3.048907897157133E-4, 1.6671472561019247E-4, &nbsp; 
<br />&gt; &gt; 4.4500080981934345E-4, -5.389129745055723E-4, &nbsp; 
<br />&gt; &gt; -2.6090705011393183E-4, -1.2633284900760366E-4, &nbsp; 
<br />&gt; &gt; -6.485428846447651E-7}, 
<br />&gt; &gt; {1.6671472561019247E-4, 2.341985572849691E-4, &nbsp; 
<br />&gt; &gt; 2.9113916450678265E-4, -4.5760873539850514E-4, &nbsp; 
<br />&gt; &gt; 1.3078636134987255E-5, -2.354289901013046E-5, -7.578030042426654E-7}, 
<br />&gt; &gt; {4.4500080981934345E-4, 2.9113916450678265E-4, 7.46023915996829E-4, &nbsp; 
<br />&gt; &gt; -0.0010247176498305568, -2.6745504327902895E-4, &nbsp; 
<br />&gt; &gt; -1.6563544154823496E-4, -8.293698990696063E-7}, 
<br />&gt; &gt; {-5.389129745055723E-4, -4.5760873539850514E-4, &nbsp; 
<br />&gt; &gt; -0.0010247176498305568, 0.001754169535149865, &nbsp; 
<br />&gt; &gt; 2.0293065310212377E-4, 2.1401092557826588E-4, 1.0252846778608953E-7}, 
<br />&gt; &gt; {-2.6090705011393183E-4, 1.3078636134987255E-5, &nbsp; 
<br />&gt; &gt; -2.6745504327902895E-4, 2.0293065310212377E-4, &nbsp; 
<br />&gt; &gt; 4.632320892679136E-4, 1.7969731066037214E-4, 2.4953495129362833E-8}, 
<br />&gt; &gt; {-1.2633284900760366E-4, -2.354289901013046E-5, &nbsp; 
<br />&gt; &gt; -1.6563544154823496E-4, 2.1401092557826588E-4, &nbsp; 
<br />&gt; &gt; 1.7969731066037214E-4, 8..346410612364995E-5, -7.02099350897589E-8}, 
<br />&gt; &gt; {-6.485428846447651E-7, -7.578030042426654E-7, &nbsp; 
<br />&gt; &gt; -8.293698990696063E-7, 1.0252846778608953E-7, &nbsp; 
<br />&gt; &gt; 2.4953495129362833E-8, -7.02099350897589E-8, 8.367244992498656E-9}} 
<br />&gt; &gt; 
<br />&gt; &gt; [c]={{0.010638291263564232}, 
<br />&gt; &gt; {0.013500370827906071}, 
<br />&gt; &gt; {0.011390037735101773}, 
<br />&gt; &gt; {0.010385042339767682}, 
<br />&gt; &gt; {3.812208389845893E-4}, 
<br />&gt; &gt; {0.002315505853720011}, 
<br />&gt; &gt; {0.0}} 
<br />&gt; &gt; 
<br />&gt; &gt; 
<br />&gt; &gt; [AI]={{1.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp; 0.0}, 
<br />&gt; &gt; {0.0, &nbsp; &nbsp; 1.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp;0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp; 0.0}, 
<br />&gt; &gt; {0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 1.0, &nbsp; &nbsp; &nbsp;0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp; 0.0}, 
<br />&gt; &gt; {0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp;1.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp; 0.0}, 
<br />&gt; &gt; {0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp;0.0, &nbsp; &nbsp; 1.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp; 0.0}, 
<br />&gt; &gt; {0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp;0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 1.0, &nbsp; &nbsp; &nbsp; 0.0}, 
<br />&gt; &gt; {0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp;0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp; 1.0}, 
<br />&gt; &gt; {-1.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp;0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp; 0.0}, 
<br />&gt; &gt; {0.0, &nbsp; &nbsp; -1.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp;0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp; 0.0}, 
<br />&gt; &gt; {0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; -1.0, &nbsp; &nbsp; &nbsp;0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp; 0.0}, 
<br />&gt; &gt; {0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp;-1.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp; 0.0}, 
<br />&gt; &gt; {0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp;0.0, &nbsp; &nbsp; -1.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp; 0.0}, 
<br />&gt; &gt; {0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp;0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; -1.0, &nbsp; &nbsp; &nbsp; 0.0}, 
<br />&gt; &gt; {0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp;0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; 0.0, &nbsp; &nbsp; &nbsp; -1.0}} 
<br />&gt; &gt; 
<br />&gt; &gt; [bi]={{1.0}, 
<br />&gt; &gt; {1.0}, 
<br />&gt; &gt; {1.0}, 
<br />&gt; &gt; {1.0}, 
<br />&gt; &gt; {1.0}, 
<br />&gt; &gt; {1.0}, 
<br />&gt; &gt; {1.0}, 
<br />&gt; &gt; {0.0}, 
<br />&gt; &gt; {0.0}, 
<br />&gt; &gt; {0.0}, 
<br />&gt; &gt; {0.0}, 
<br />&gt; &gt; {0.0}, 
<br />&gt; &gt; {0.0}, 
<br />&gt; &gt; {0.0}} 
<br />&gt; &gt; 
<br />&gt; &gt; And here's the result: 
<br />&gt; &gt; 
<br />&gt; &gt; -0.25 
<br />&gt; &gt; -0.25 
<br />&gt; &gt; -0.25 
<br />&gt; &gt; -0.25 
<br />&gt; &gt; 0.75 
<br />&gt; &gt; 0.75 
<br />&gt; &gt; 0.75 
<br />&gt; &gt; 
<br />&gt; &gt; And i found that if i change the nagetive ExceptedReturn to &nbsp; 
<br />&gt; &gt; positive, the result will be fine. 
<br />&gt; &gt; 
<br />&gt; &gt; 
<br />&gt; &gt; On Wed, 14 Jan 2009 09:27:53 +0100, Anders Peterson wrote 
<br />&gt; &gt; &gt; How small are the negative values? 
<br />&gt; &gt; &gt; 
<br />&gt; &gt; &gt; Give me a unit test (something that I can easily integrate with the 
<br />&gt; &gt; &gt; existing junit tests) that shows what you do, and where/how you 
<br />&gt; &gt; &gt; think &nbsp;it goes wrong. 
<br />&gt; &gt; &gt; 
<br />&gt; &gt; &gt; /Anders 
<br />&gt; &gt; &gt; 
<br />&gt; &gt; &gt; On 14 jan 2009, at 03.58, Ruan Weichao wrote: 
<br />&gt; &gt; &gt; 
<br />&gt; &gt; &gt; &gt; 
<br />&gt; &gt; &gt; &gt; Hi, 
<br />&gt; &gt; &gt; &gt; 
<br />&gt; &gt; &gt; &gt; I use the &nbsp;Markowitz model to calculate the optimize some assets' 
<br />&gt; &gt; &gt; &gt; weights, 
<br />&gt; &gt; &gt; &gt; 
<br />&gt; &gt; &gt; &gt; and set all the assets' lowerlimit with Zero and set all the 
<br />&gt; &gt; &gt; &gt; upperlimits with 
<br />&gt; &gt; &gt; &gt; 
<br />&gt; &gt; &gt; &gt; One, but it turns out with some negative value which are little 
<br />&gt; &gt; &gt; &gt; than Zero, it 
<br />&gt; &gt; &gt; &gt; 
<br />&gt; &gt; &gt; &gt; seems that the lowerlimit does not work. 
<br />&gt; &gt; &gt; &gt; 
<br />&gt; &gt; &gt; &gt; So I want to ask what should I do it to confirm the lowerlimit? 
<br />&gt; &gt; &gt; &gt; 
<br />&gt; &gt; &gt; &gt; -- 
<br />&gt; &gt; &gt; &gt; Best regards 
<br />&gt; &gt; &gt; &gt; 
<br />&gt; &gt; &gt; &gt; weichao 
<br />&gt; &gt; &gt; &gt; 
<br />&gt; &gt; &gt; &gt; &nbsp; 
<br />&gt; &gt; ---------------------------------------------------------------------- 
<br />&gt; &gt; &gt; &gt; -------- 
<br />&gt; &gt; &gt; &gt; This SF.net email is sponsored by: 
<br />&gt; &gt; &gt; &gt; SourcForge Community 
<br />&gt; &gt; &gt; &gt; SourceForge wants to tell your story. 
<br />&gt; &gt; &gt; &gt; <a href="http://p.sf.net/sfu/sf-spreadtheword" target="_blank">http://p.sf.net/sfu/sf-spreadtheword</a> 
<br />&gt; &gt; &gt; &gt; _______________________________________________ 
<br />&gt; &gt; &gt; &gt; ojAlgo-user mailing list 
<br />&gt; &gt; &gt; &gt; ojA...@li... 
<br />&gt; &gt; &gt; &gt; <a href="https://lists.sourceforge.net/lists/listinfo/ojalgo-user" target="_blank">https://lists.sourceforge.net/lists/listinfo/ojalgo-user</a> 
<br />&gt; &gt; &gt; &gt; 
<br />&gt; &gt; &gt; &gt; 
<br />&gt; &gt; &gt; 
<br />&gt; &gt; &gt; &nbsp; 
<br />&gt; &gt; ---------------------------------------------------------------------- 
<br />&gt; &gt; ------ 
<br />&gt; &gt; -- 
<br />&gt; &gt; &gt; This SF.net email is sponsored by: 
<br />&gt; &gt; &gt; SourcForge Community 
<br />&gt; &gt; &gt; SourceForge wants to tell your story. 
<br />&gt; &gt; &gt; <a href="http://p.sf.net/sfu/sf-spreadtheword" target="_blank">http://p.sf.net/sfu/sf-spreadtheword</a> 
<br />&gt; &gt; &gt; _______________________________________________ 
<br />&gt; &gt; &gt; ojAlgo-user mailing list 
<br />&gt; &gt; &gt; ojA...@li... 
<br />&gt; &gt; &gt; <a href="https://lists.sourceforge.net/lists/listinfo/ojalgo-user" target="_blank">https://lists.sourceforge.net/lists/listinfo/ojalgo-user</a> 
<br />&gt; &gt; 
<br />&gt; &gt; 
<br />&gt; &gt; -- 
<br />&gt; &gt; Best regards 
<br />&gt; &gt; 
<br />&gt; &gt; ---------------------------------------------------------------------- 
<br />&gt; &gt; -------- 
<br />&gt; &gt; This SF.net email is sponsored by: 
<br />&gt; &gt; SourcForge Community 
<br />&gt; &gt; SourceForge wants to tell your story. 
<br />&gt; &gt; <a href="http://p.sf.net/sfu/sf" target="_blank">http://p.sf.net/sfu/sf</a>- 
<br />&gt; &gt; spreadtheword_______________________________________________ 
<br />&gt; &gt; ojAlgo-user mailing list 
<br />&gt; &gt; ojA...@li... 
<br />&gt; &gt; <a href="https://lists.sourceforge.net/lists/listinfo/ojalgo-user" target="_blank">https://lists.sourceforge.net/lists/listinfo/ojalgo-user</a> 
<br />&gt; 
<br />&gt; ---------------------------------------------------------------------------- 
<br />-- 
<br />&gt; This SF.net email is sponsored by: 
<br />&gt; SourcForge Community 
<br />&gt; SourceForge wants to tell your story. 
<br />&gt; <a href="http://p.sf.net/sfu/sf-spreadtheword" target="_blank">http://p.sf.net/sfu/sf-spreadtheword</a> 
<br />&gt; _______________________________________________ 
<br />&gt; ojAlgo-user mailing list 
<br />&gt; ojA...@li... 
<br />&gt; <a href="https://lists.sourceforge.net/lists/listinfo/ojalgo-user" target="_blank">https://lists.sourceforge.net/lists/listinfo/ojalgo-user</a> 
<br />
<br />
<br />-- 
<br />Best regards 
<br /></font></p>

</BODY>
</HTML>

Re: [ojAlgo-user] Markowitz Model Problem

[Anders P. <an...@op...>]

Your expected excess returns matrix looks like this:

class org.ojalgo.matrix.BigMatrix
{{-0.010638291263564232},
{-0.013500370827906071},
{-0.011390037735101773},
{-0.010385042339767682},
{-0.0003812208389845893},
{-0.002315505853720011},
{0}}

Is that correct?

Earlier you wrote "And i found that if i change the nagetive  
ExceptedReturn to positive, the result will be fine."

/Anders


On 15 jan 2009, at 11.10, Ruan Weichao wrote:

> Hi，i am using v25 and here's my code:
>
> / 
> ********************************************************************** 
> **********/
>
> import java.math.BigDecimal;
> import java.util.List;
>
> import org.ojalgo.finance.portfolio.MarketEquilibrium;
> import org.ojalgo.finance.portfolio.MarkowitzModel;
> import org.ojalgo.matrix.*;
> import org.ojalgo.matrix.store.BigDenseStore;
> import org.ojalgo.matrix.store.MatrixStore;
>
>
> public class TestMarkowitz {
>
>  public double getCovarance(double[] value1, double[] value2){
>
>   int n = value1.length;
>   double averOne = 0;
>   double averTwo = 0;
>   for(int i=0;i<n;i++)
>   {
>    averOne+=value1[i];
>    averTwo+=value2[i];
>   }
>   averOne /= n;
>   averTwo /= n;
>
>   double sum = 0;
>   for(int i=0;i<n;i++)
>    sum+=(value1[i]-averOne)*(value2[i]-averTwo);
>
>   return sum / (n-1);
>  }
>
>  public BasicMatrix getACovariances(double[][] returns){
>
>   int row = returns.length;
>   int col = returns[0].length;
>   MatrixStore<BigDecimal> a = BigDenseStore.FACTORY..makeZero(row,  
> col);
>   BasicMatrix covariances = new BigMatrix(a);
>
>   for(int i=1;i<=row;i++){
>    for(int j=i;j<=col;j++){
>     double tmp = this.getCovarance(returns[i-1], returns[j-1]);
>     covariances = covariances.set(i-1, j-1,tmp);
>     covariances = covariances.set(j-1, i-1,tmp);
>    }
>   }
>   return covariances;
>  }
>
>  public BasicMatrix getAnExpectedExcessReturns(double[][] returns){
>
>   int row = returns.length;
>   int col = returns[0].length;
>
>   MatrixStore<BigDecimal> a = BigDenseStore.FACTORY..makeZero(row, 1);
>   BasicMatrix expectedExcessReturns = new BigMatrix(a);
>
>   double riskFreeReturn = 0;
>   for(int i=1;i<=col;i++)
>    riskFreeReturn+=returns[row-1][i-1];
>   riskFreeReturn/=row;
>
>   for(int i=1;i<=row;i++){
>
>    double tmp = 0;
>    for(int j=1;j<=col;j++)
>     tmp+=returns[i-1][j-1];
>    tmp/=col;
>
>    expectedExcessReturns = expectedExcessReturns.set(i-1, 0, tmp- 
> riskFreeReturn);
>   }
>
>   return expectedExcessReturns;
>  }
>
>  public static void main(String[] args)
>  {
>   int assetNum=7;
>   double[][] assets_return={
>    {-1.5905837442343828E-4, -0.03062360801781757,  
> -0.029857534032853142, -0.011811692726036832,  
> -0.017972310602803136, 0.017338003502626997, 0.0},
>    {-0.02757158006362653, -0.02562704471101405,  
> -0.011751538891997735, -0.024915062287655786, -0.01684088269454123,  
> 0.013585351447135364, 0.0},
>    {-0.00699300699300693, -0.033802816901408676,  
> -0.04675196850393671, -0.021166752710376546, -0.007911392405063583,  
> 0.03827751196172254, 0.0},
>    {-0.007626310772164015, 0.0038424591738713027,  
> 0.02488038277511978, 0.025210084033613675, -0.02003642987249557,  
> -0.09758364312267642, 0.0},
>    {-0.03965053763440893, 0.021693491952414375,  
> 0.01643835616438392, -0.007412398921833087, 0.01765105227427014,  
> -0.010006671114076025, 0.0},
>    {-0.017821782178217872, 0.005040322580645311,  
> 0.006018054162487363, 9.008107296569024E-4, 0.002999999999999824,  
> -0.01196410767696908, 0.0},
>    {2.630552127527583E-4, 2.5867028174649627E-4,  
> 2.3866431891514327E-4, 1.9564035993080523E-4, 2.351016690966669E-4,  
> 1.9070675120065465E-4, 0.0}
>   };
>
>   TestMarkowitz tm = new TestMarkowitz();
>   BasicMatrix covariances = tm.getACovariances(assets_return);
>   BasicMatrix expectedExcessReturns = tm.getAnExpectedExcessReturns 
> (assets_return);
>   BigDecimal riskAversion = new BigDecimal(1.0);
>
>   MarketEquilibrium marketEquilibrium = new MarketEquilibrium 
> (covariances,riskAversion);
>   MarkowitzModel markowitzModel = new MarkowitzModel 
> (marketEquilibrium,expectedExcessReturns);
>
>   for(int i=0;i<assetNum;i++){
>    markowitzModel.setLowerLimit(i, new BigDecimal(0.0));
>    markowitzModel.setUpperLimit(i, new BigDecimal(1.0));
>   }
>   List<BigDecimal> re = markowitzModel.getWeights();
>
>   System.out.println("=======result====================");
>   for(int i=0;i<re.size();i++)
>    System.out.println(re.get(i));
>
>   return;
>  }
>
> }
>
> / 
> ********************************************************************** 
> **********/
>
> and the result turn out as:
>
> -0.25
> -0.25
> -0.25
> -0.25
> 0.75
> 0.75
> 0.75
>
> / 
> ********************************************************************** 
> ******/
>
> If you find out any problem, just let me know , thanks a lot!
>
> On Thu, 15 Jan 2009 09:20:52 +0100, Anders Peterson wrote
> > You have to send code (snippets) that I can execute.
> >
> > Are you using v25 or the latest from cvs?
> >
> > How do you instantiate the ActiveSetSolver?
> >
> > What's the correct solution?
> >
> > You say that if you change the sign of "C" you get an ok result.
> > What  does this mean? Did you make a mistake with the sign, or do
> > you  suspect there is a bug in ojAlgo?
> >
> > I need better info regarding what you do, and what the problem is!
> >
> > /Anders
> >
> > On 15 jan 2009, at 06.14, Ruan Weichao wrote:
> >
> > >
> > > Thank you for your response.
> > >
> > > Here's my model's detail:
> > >
> > > [AE]={1.0,     1.0,     1.0,       1.0,     1.0,     1.0,        
> 1.0}
> > >
> > >
> > > [be]={1.0}
> > >
> > >
> > > [Q]={3.048907897157133E-4, 1.6671472561019247E-4,
> > > 4.4500080981934345E-4, -5.389129745055723E-4,
> > > -2.6090705011393183E-4, -1.2633284900760366E-4,
> > > -6.485428846447651E-7},
> > > {1.6671472561019247E-4, 2.341985572849691E-4,
> > > 2.9113916450678265E-4, -4.5760873539850514E-4,
> > > 1.3078636134987255E-5, -2.354289901013046E-5,  
> -7.578030042426654E-7},
> > > {4.4500080981934345E-4, 2.9113916450678265E-4,  
> 7.46023915996829E-4,
> > > -0.0010247176498305568, -2.6745504327902895E-4,
> > > -1.6563544154823496E-4, -8.293698990696063E-7},
> > > {-5.389129745055723E-4, -4.5760873539850514E-4,
> > > -0.0010247176498305568, 0.001754169535149865,
> > > 2.0293065310212377E-4, 2.1401092557826588E-4,  
> 1.0252846778608953E-7},
> > > {-2.6090705011393183E-4, 1.3078636134987255E-5,
> > > -2.6745504327902895E-4, 2.0293065310212377E-4,
> > > 4.632320892679136E-4, 1.7969731066037214E-4,  
> 2.4953495129362833E-8},
> > > {-1.2633284900760366E-4, -2.354289901013046E-5,
> > > -1.6563544154823496E-4, 2.1401092557826588E-4,
> > > 1.7969731066037214E-4, 8..346410612364995E-5,  
> -7.02099350897589E-8},
> > > {-6.485428846447651E-7, -7.578030042426654E-7,
> > > -8.293698990696063E-7, 1.0252846778608953E-7,
> > > 2.4953495129362833E-8, -7.02099350897589E-8,  
> 8.367244992498656E-9}}
> > >
> > > [c]={{0.010638291263564232},
> > > {0.013500370827906071},
> > > {0.011390037735101773},
> > > {0.010385042339767682},
> > > {3.812208389845893E-4},
> > > {0.002315505853720011},
> > > {0.0}}
> > >
> > >
> > > [AI]={{1.0,     0.0,     0.0,       0.0,     0.0,      
> 0.0,       0.0},
> > > {0.0,     1.0,     0.0,      0.0,     0.0,     0.0,       0.0},
> > > {0.0,     0.0,     1.0,      0.0,     0.0,     0.0,       0.0},
> > > {0.0,     0.0,     0.0,      1.0,     0.0,     0.0,       0.0},
> > > {0.0,     0.0,     0.0,      0.0,     1.0,     0.0,       0.0},
> > > {0.0,     0.0,     0.0,      0.0,     0.0,     1.0,       0.0},
> > > {0.0,     0.0,     0.0,      0.0,     0.0,     0.0,       1.0},
> > > {-1.0,     0.0,     0.0,      0.0,     0.0,     0.0,       0.0},
> > > {0.0,     -1.0,     0.0,      0.0,     0.0,     0.0,       0.0},
> > > {0.0,     0.0,     -1.0,      0.0,     0.0,     0.0,       0.0},
> > > {0.0,     0.0,     0.0,      -1.0,     0.0,     0.0,       0.0},
> > > {0.0,     0.0,     0.0,      0.0,     -1.0,     0.0,       0.0},
> > > {0.0,     0.0,     0.0,      0.0,     0.0,     -1.0,       0.0},
> > > {0.0,     0.0,     0.0,      0.0,     0.0,     0.0,       -1.0}}
> > >
> > > [bi]={{1.0},
> > > {1.0},
> > > {1.0},
> > > {1.0},
> > > {1.0},
> > > {1.0},
> > > {1.0},
> > > {0.0},
> > > {0.0},
> > > {0.0},
> > > {0.0},
> > > {0.0},
> > > {0.0},
> > > {0.0}}
> > >
> > > And here's the result:
> > >
> > > -0.25
> > > -0.25
> > > -0.25
> > > -0.25
> > > 0.75
> > > 0.75
> > > 0.75
> > >
> > > And i found that if i change the nagetive ExceptedReturn to
> > > positive, the result will be fine.
> > >
> > >
> > > On Wed, 14 Jan 2009 09:27:53 +0100, Anders Peterson wrote
> > > > How small are the negative values?
> > > >
> > > > Give me a unit test (something that I can easily integrate  
> with the
> > > > existing junit tests) that shows what you do, and where/how you
> > > > think  it goes wrong.
> > > >
> > > > /Anders
> > > >
> > > > On 14 jan 2009, at 03.58, Ruan Weichao wrote:
> > > >
> > > > >
> > > > > Hi,
> > > > >
> > > > > I use the  Markowitz model to calculate the optimize some  
> assets'
> > > > > weights,
> > > > >
> > > > > and set all the assets' lowerlimit with Zero and set all the
> > > > > upperlimits with
> > > > >
> > > > > One, but it turns out with some negative value which are  
> little
> > > > > than Zero, it
> > > > >
> > > > > seems that the lowerlimit does not work.
> > > > >
> > > > > So I want to ask what should I do it to confirm the  
> lowerlimit?
> > > > >
> > > > > --
> > > > > Best regards
> > > > >
> > > > > weichao
> > > > >
> > > > >
> > >  
> ----------------------------------------------------------------------
> > > > > --------
> > > > > This SF.net email is sponsored by:
> > > > > SourcForge Community
> > > > > SourceForge wants to tell your story.
> > > > > http://p.sf.net/sfu/sf-spreadtheword
> > > > > _______________________________________________
> > > > > ojAlgo-user mailing list
> > > > > ojA...@li...
> > > > > https://lists.sourceforge.net/lists/listinfo/ojalgo-user
> > > > >
> > > > >
> > > >
> > > >
> > >  
> ----------------------------------------------------------------------
> > > ------
> > > --
> > > > This SF.net email is sponsored by:
> > > > SourcForge Community
> > > > SourceForge wants to tell your story.
> > > > http://p.sf.net/sfu/sf-spreadtheword
> > > > _______________________________________________
> > > > ojAlgo-user mailing list
> > > > ojA...@li...
> > > > https://lists.sourceforge.net/lists/listinfo/ojalgo-user
> > >
> > >
> > > --
> > > Best regards
> > >
> > >  
> ----------------------------------------------------------------------
> > > --------
> > > This SF.net email is sponsored by:
> > > SourcForge Community
> > > SourceForge wants to tell your story.
> > > http://p.sf.net/sfu/sf-
> > > spreadtheword_______________________________________________
> > > ojAlgo-user mailing list
> > > ojA...@li...
> > > https://lists.sourceforge.net/lists/listinfo/ojalgo-user
> >
> >  
> ---------------------------------------------------------------------- 
> ------
> -- 
> > This SF.net email is sponsored by:
> > SourcForge Community
> > SourceForge wants to tell your story.
> > http://p.sf.net/sfu/sf-spreadtheword
> > _______________________________________________
> > ojAlgo-user mailing list
> > ojA...@li...
> > https://lists.sourceforge.net/lists/listinfo/ojalgo-user
>
>
> -- 
> Best regards
>
> ---------------------------------------------------------------------- 
> --------
> This SF.net email is sponsored by:
> SourcForge Community
> SourceForge wants to tell your story.
> http://p.sf.net/sfu/sf- 
> spreadtheword_______________________________________________
> ojAlgo-user mailing list
> ojA...@li...
> https://lists.sourceforge.net/lists/listinfo/ojalgo-user

Re: [ojAlgo-user] Markowitz Model Problem

[Ruan W. <is0...@ma...>]

Hi,

Yes, that's the expected excess returns of my program.

Is it right?

Can the "expected excess returns" be negative?

On Thu, 15 Jan 2009 14:09:43 +0100, Anders Peterson wrote
> Your expected excess returns matrix looks like this:
> 
> class org.ojalgo.matrix.BigMatrix
> {{-0.010638291263564232},
> {-0.013500370827906071},
> {-0.011390037735101773},
> {-0.010385042339767682},
> {-0.0003812208389845893},
> {-0.002315505853720011},
> {0}}
> 
> Is that correct?
> 
> Earlier you wrote "And i found that if i change the nagetive  
> ExceptedReturn to positive, the result will be fine."
> 
> /Anders
> 
> On 15 jan 2009, at 11.10, Ruan Weichao wrote:
> 
> > Hi，i am using v25 and here's my code:
> >
> > / 
> > ********************************************************************** 
> > **********/
> >
> > import java.math.BigDecimal;
> > import java.util.List;
> >
> > import org.ojalgo.finance.portfolio.MarketEquilibrium;
> > import org.ojalgo.finance.portfolio.MarkowitzModel;
> > import org.ojalgo.matrix.*;
> > import org.ojalgo.matrix.store.BigDenseStore;
> > import org.ojalgo.matrix.store.MatrixStore;
> >
> >
> > public class TestMarkowitz {
> >
> >  public double getCovarance(double[] value1, double[] value2){
> >
> >   int n = value1.length;
> >   double averOne = 0;
> >   double averTwo = 0;
> >   for(int i=0;i<n;i++)
> >   {
> >    averOne+=value1[i];
> >    averTwo+=value2[i];
> >   }
> >   averOne /= n;
> >   averTwo /= n;
> >
> >   double sum = 0;
> >   for(int i=0;i<n;i++)
> >    sum+=(value1[i]-averOne)*(value2[i]-averTwo);
> >
> >   return sum / (n-1);
> >  }
> >
> >  public BasicMatrix getACovariances(double[][] returns){
> >
> >   int row = returns.length;
> >   int col = returns[0].length;
> >   MatrixStore<BigDecimal> a = BigDenseStore.FACTORY..makeZero(row,  
> > col);
> >   BasicMatrix covariances = new BigMatrix(a);
> >
> >   for(int i=1;i<=row;i++){
> >    for(int j=i;j<=col;j++){
> >     double tmp = this.getCovarance(returns[i-1], returns[j-1]);
> >     covariances = covariances.set(i-1, j-1,tmp);
> >     covariances = covariances.set(j-1, i-1,tmp);
> >    }
> >   }
> >   return covariances;
> >  }
> >
> >  public BasicMatrix getAnExpectedExcessReturns(double[][] returns){
> >
> >   int row = returns.length;
> >   int col = returns[0].length;
> >
> >   MatrixStore<BigDecimal> a = BigDenseStore.FACTORY..makeZero(row, 1);
> >   BasicMatrix expectedExcessReturns = new BigMatrix(a);
> >
> >   double riskFreeReturn = 0;
> >   for(int i=1;i<=col;i++)
> >    riskFreeReturn+=returns[row-1][i-1];
> >   riskFreeReturn/=row;
> >
> >   for(int i=1;i<=row;i++){
> >
> >    double tmp = 0;
> >    for(int j=1;j<=col;j++)
> >     tmp+=returns[i-1][j-1];
> >    tmp/=col;
> >
> >    expectedExcessReturns = expectedExcessReturns.set(i-1, 0, tmp- 
> > riskFreeReturn);
> >   }
> >
> >   return expectedExcessReturns;
> >  }
> >
> >  public static void main(String[] args)
> >  {
> >   int assetNum=7;
> >   double[][] assets_return={
> >    {-1.5905837442343828E-4, -0.03062360801781757,  
> > -0.029857534032853142, -0.011811692726036832,  
> > -0.017972310602803136, 0.017338003502626997, 0.0},
> >    {-0.02757158006362653, -0.02562704471101405,  
> > -0.011751538891997735, -0.024915062287655786, -0.01684088269454123,  
> > 0.013585351447135364, 0.0},
> >    {-0.00699300699300693, -0.033802816901408676,  
> > -0.04675196850393671, -0.021166752710376546, -0.007911392405063583,  
> > 0.03827751196172254, 0.0},
> >    {-0.007626310772164015, 0.0038424591738713027,  
> > 0.02488038277511978, 0.025210084033613675, -0.02003642987249557,  
> > -0.09758364312267642, 0.0},
> >    {-0.03965053763440893, 0.021693491952414375,  
> > 0.01643835616438392, -0.007412398921833087, 0.01765105227427014,  
> > -0.010006671114076025, 0.0},
> >    {-0.017821782178217872, 0.005040322580645311,  
> > 0.006018054162487363, 9.008107296569024E-4, 0.002999999999999824,  
> > -0.01196410767696908, 0.0},
> >    {2.630552127527583E-4, 2.5867028174649627E-4,  
> > 2.3866431891514327E-4, 1.9564035993080523E-4, 2.351016690966669E-4,  
> > 1.9070675120065465E-4, 0.0}
> >   };
> >
> >   TestMarkowitz tm = new TestMarkowitz();
> >   BasicMatrix covariances = tm.getACovariances(assets_return);
> >   BasicMatrix expectedExcessReturns = tm.getAnExpectedExcessReturns 
> > (assets_return);
> >   BigDecimal riskAversion = new BigDecimal(1.0);
> >
> >   MarketEquilibrium marketEquilibrium = new MarketEquilibrium 
> > (covariances,riskAversion);
> >   MarkowitzModel markowitzModel = new MarkowitzModel 
> > (marketEquilibrium,expectedExcessReturns);
> >
> >   for(int i=0;i<assetNum;i++){
> >    markowitzModel.setLowerLimit(i, new BigDecimal(0.0));
> >    markowitzModel.setUpperLimit(i, new BigDecimal(1.0));
> >   }
> >   List<BigDecimal> re = markowitzModel.getWeights();
> >
> >   System.out.println("=======result====================");
> >   for(int i=0;i<re.size();i++)
> >    System.out.println(re.get(i));
> >
> >   return;
> >  }
> >
> > }
> >
> > / 
> > ********************************************************************** 
> > **********/
> >
> > and the result turn out as:
> >
> > -0.25
> > -0.25
> > -0.25
> > -0.25
> > 0.75
> > 0.75
> > 0.75
> >
> > / 
> > ********************************************************************** 
> > ******/
> >
> > If you find out any problem, just let me know , thanks a lot!
> >
> > On Thu, 15 Jan 2009 09:20:52 +0100, Anders Peterson wrote
> > > You have to send code (snippets) that I can execute.
> > >
> > > Are you using v25 or the latest from cvs?
> > >
> > > How do you instantiate the ActiveSetSolver?
> > >
> > > What's the correct solution?
> > >
> > > You say that if you change the sign of "C" you get an ok result.
> > > What  does this mean? Did you make a mistake with the sign, or do
> > > you  suspect there is a bug in ojAlgo?
> > >
> > > I need better info regarding what you do, and what the problem is!
> > >
> > > /Anders
> > >
> > > On 15 jan 2009, at 06.14, Ruan Weichao wrote:
> > >
> > > >
> > > > Thank you for your response.
> > > >
> > > > Here's my model's detail:
> > > >
> > > > [AE]={1.0,     1.0,     1.0,       1.0,     1.0,     1.0,        
> > 1.0}
> > > >
> > > >
> > > > [be]={1.0}
> > > >
> > > >
> > > > [Q]={3.048907897157133E-4, 1.6671472561019247E-4,
> > > > 4.4500080981934345E-4, -5.389129745055723E-4,
> > > > -2.6090705011393183E-4, -1.2633284900760366E-4,
> > > > -6.485428846447651E-7},
> > > > {1.6671472561019247E-4, 2.341985572849691E-4,
> > > > 2.9113916450678265E-4, -4.5760873539850514E-4,
> > > > 1.3078636134987255E-5, -2.354289901013046E-5,  
> > -7.578030042426654E-7},
> > > > {4.4500080981934345E-4, 2.9113916450678265E-4,  
> > 7.46023915996829E-4,
> > > > -0.0010247176498305568, -2.6745504327902895E-4,
> > > > -1.6563544154823496E-4, -8.293698990696063E-7},
> > > > {-5.389129745055723E-4, -4.5760873539850514E-4,
> > > > -0.0010247176498305568, 0.001754169535149865,
> > > > 2.0293065310212377E-4, 2.1401092557826588E-4,  
> > 1.0252846778608953E-7},
> > > > {-2.6090705011393183E-4, 1.3078636134987255E-5,
> > > > -2.6745504327902895E-4, 2.0293065310212377E-4,
> > > > 4.632320892679136E-4, 1.7969731066037214E-4,  
> > 2.4953495129362833E-8},
> > > > {-1.2633284900760366E-4, -2.354289901013046E-5,
> > > > -1.6563544154823496E-4, 2.1401092557826588E-4,
> > > > 1.7969731066037214E-4, 8..346410612364995E-5,  
> > -7.02099350897589E-8},
> > > > {-6.485428846447651E-7, -7.578030042426654E-7,
> > > > -8.293698990696063E-7, 1.0252846778608953E-7,
> > > > 2.4953495129362833E-8, -7.02099350897589E-8,  
> > 8.367244992498656E-9}}
> > > >
> > > > [c]={{0.010638291263564232},
> > > > {0.013500370827906071},
> > > > {0.011390037735101773},
> > > > {0.010385042339767682},
> > > > {3.812208389845893E-4},
> > > > {0.002315505853720011},
> > > > {0.0}}
> > > >
> > > >
> > > > [AI]={{1.0,     0.0,     0.0,       0.0,     0.0,      
> > 0.0,       0.0},
> > > > {0.0,     1.0,     0.0,      0.0,     0.0,     0.0,       0.0},
> > > > {0.0,     0.0,     1.0,      0.0,     0.0,     0.0,       0.0},
> > > > {0.0,     0.0,     0.0,      1.0,     0.0,     0.0,       0.0},
> > > > {0.0,     0.0,     0.0,      0.0,     1.0,     0.0,       0.0},
> > > > {0.0,     0.0,     0.0,      0.0,     0.0,     1.0,       0.0},
> > > > {0.0,     0.0,     0.0,      0.0,     0.0,     0.0,       1.0},
> > > > {-1.0,     0.0,     0.0,      0.0,     0.0,     0.0,       0.0},
> > > > {0.0,     -1.0,     0.0,      0.0,     0.0,     0.0,       0.0},
> > > > {0.0,     0.0,     -1.0,      0.0,     0.0,     0.0,       0.0},
> > > > {0.0,     0.0,     0.0,      -1.0,     0.0,     0.0,       0.0},
> > > > {0.0,     0.0,     0.0,      0.0,     -1.0,     0.0,       0.0},
> > > > {0.0,     0.0,     0.0,      0.0,     0.0,     -1.0,       0.0},
> > > > {0.0,     0.0,     0.0,      0.0,     0.0,     0.0,       -1.0}}
> > > >
> > > > [bi]={{1.0},
> > > > {1.0},
> > > > {1.0},
> > > > {1.0},
> > > > {1.0},
> > > > {1.0},
> > > > {1.0},
> > > > {0.0},
> > > > {0.0},
> > > > {0.0},
> > > > {0.0},
> > > > {0.0},
> > > > {0.0},
> > > > {0.0}}
> > > >
> > > > And here's the result:
> > > >
> > > > -0.25
> > > > -0.25
> > > > -0.25
> > > > -0.25
> > > > 0.75
> > > > 0.75
> > > > 0.75
> > > >
> > > > And i found that if i change the nagetive ExceptedReturn to
> > > > positive, the result will be fine.
> > > >
> > > >
> > > > On Wed, 14 Jan 2009 09:27:53 +0100, Anders Peterson wrote
> > > > > How small are the negative values?
> > > > >
> > > > > Give me a unit test (something that I can easily integrate  
> > with the
> > > > > existing junit tests) that shows what you do, and where/how you
> > > > > think  it goes wrong.
> > > > >
> > > > > /Anders
> > > > >
> > > > > On 14 jan 2009, at 03.58, Ruan Weichao wrote:
> > > > >
> > > > > >
> > > > > > Hi,
> > > > > >
> > > > > > I use the  Markowitz model to calculate the optimize some  
> > assets'
> > > > > > weights,
> > > > > >
> > > > > > and set all the assets' lowerlimit with Zero and set all the
> > > > > > upperlimits with
> > > > > >
> > > > > > One, but it turns out with some negative value which are  
> > little
> > > > > > than Zero, it
> > > > > >
> > > > > > seems that the lowerlimit does not work.
> > > > > >
> > > > > > So I want to ask what should I do it to confirm the  
> > lowerlimit?
> > > > > >
> > > > > > --
> > > > > > Best regards
> > > > > >
> > > > > > weichao
> > > > > >
> > > > > >
> > > >  
> > ----------------------------------------------------------------------
> > > > > > --------
> > > > > > This SF.net email is sponsored by:
> > > > > > SourcForge Community
> > > > > > SourceForge wants to tell your story.
> > > > > > http://p.sf.net/sfu/sf-spreadtheword
> > > > > > _______________________________________________
> > > > > > ojAlgo-user mailing list
> > > > > > ojA...@li...
> > > > > > https://lists.sourceforge.net/lists/listinfo/ojalgo-user
> > > > > >
> > > > > >
> > > > >
> > > > >
> > > >  
> > ----------------------------------------------------------------------
> > > > ------
> > > > --
> > > > > This SF.net email is sponsored by:
> > > > > SourcForge Community
> > > > > SourceForge wants to tell your story.
> > > > > http://p.sf.net/sfu/sf-spreadtheword
> > > > > _______________________________________________
> > > > > ojAlgo-user mailing list
> > > > > ojA...@li...
> > > > > https://lists.sourceforge.net/lists/listinfo/ojalgo-user
> > > >
> > > >
> > > > --
> > > > Best regards
> > > >
> > > >  
> > ----------------------------------------------------------------------
> > > > --------
> > > > This SF.net email is sponsored by:
> > > > SourcForge Community
> > > > SourceForge wants to tell your story.
> > > > http://p.sf.net/sfu/sf-
> > > > spreadtheword_______________________________________________
> > > > ojAlgo-user mailing list
> > > > ojA...@li...
> > > > https://lists.sourceforge.net/lists/listinfo/ojalgo-user
> > >
> > >  
> > ---------------------------------------------------------------------- 
> > ------
> > -- 
> > > This SF.net email is sponsored by:
> > > SourcForge Community
> > > SourceForge wants to tell your story.
> > > http://p.sf.net/sfu/sf-spreadtheword
> > > _______________________________________________
> > > ojAlgo-user mailing list
> > > ojA...@li...
> > > https://lists.sourceforge.net/lists/listinfo/ojalgo-user
> >
> >
> > -- 
> > Best regards
> >
> > ---------------------------------------------------------------------- 
> > --------
> > This SF.net email is sponsored by:
> > SourcForge Community
> > SourceForge wants to tell your story.
> > http://p.sf.net/sfu/sf- 
> > spreadtheword_______________________________________________
> > ojAlgo-user mailing list
> > ojA...@li...
> > https://lists.sourceforge.net/lists/listinfo/ojalgo-user
> 
> ----------------------------------------------------------------------------
--
> This SF.net email is sponsored by:
> SourcForge Community
> SourceForge wants to tell your story.
> http://p.sf.net/sfu/sf-spreadtheword
> _______________________________________________
> ojAlgo-user mailing list
> ojA...@li...
> https://lists.sourceforge.net/lists/listinfo/ojalgo-user


--
Best regards

Re: [ojAlgo-user] Markowitz Model Problem

[Anders P. <an...@op...>]

On 16 jan 2009, at 02.28, Ruan Weichao wrote:

> Hi,
>
> Yes, that's the expected excess returns of my program.
>
> Is it right?
>
> Can the "expected excess returns" be negative?


I have no way of knowing what model/problem you want to solve...

In a general case excess returns (what you get for taking a risk)  
should not be negative. If they are, you do not need to optimise  
anything. Just invest everything in the risk free asset.

I'm not sure exactly why you get those negative (constraint breaking)  
portfolio weights. I'm going to look into this, but later. Don't hold  
your breath. For you, I think the solution is to work on your model.

/Anders


> On Thu, 15 Jan 2009 14:09:43 +0100, Anders Peterson wrote
>> Your expected excess returns matrix looks like this:
>>
>> class org.ojalgo.matrix.BigMatrix
>> {{-0.010638291263564232},
>> {-0.013500370827906071},
>> {-0.011390037735101773},
>> {-0.010385042339767682},
>> {-0.0003812208389845893},
>> {-0.002315505853720011},
>> {0}}
>>
>> Is that correct?
>>
>> Earlier you wrote "And i found that if i change the nagetive
>> ExceptedReturn to positive, the result will be fine."
>>
>> /Anders
>>
>> On 15 jan 2009, at 11.10, Ruan Weichao wrote:
>>
>>> Hi，i am using v25 and here's my code:
>>>
>>> /
>>> ******************************************************************** 
>>> **
>>> **********/
>>>
>>> import java.math.BigDecimal;
>>> import java.util.List;
>>>
>>> import org.ojalgo.finance.portfolio.MarketEquilibrium;
>>> import org.ojalgo.finance.portfolio.MarkowitzModel;
>>> import org.ojalgo.matrix.*;
>>> import org.ojalgo.matrix.store.BigDenseStore;
>>> import org.ojalgo.matrix.store.MatrixStore;
>>>
>>>
>>> public class TestMarkowitz {
>>>
>>>  public double getCovarance(double[] value1, double[] value2){
>>>
>>>   int n = value1.length;
>>>   double averOne = 0;
>>>   double averTwo = 0;
>>>   for(int i=0;i<n;i++)
>>>   {
>>>    averOne+=value1[i];
>>>    averTwo+=value2[i];
>>>   }
>>>   averOne /= n;
>>>   averTwo /= n;
>>>
>>>   double sum = 0;
>>>   for(int i=0;i<n;i++)
>>>    sum+=(value1[i]-averOne)*(value2[i]-averTwo);
>>>
>>>   return sum / (n-1);
>>>  }
>>>
>>>  public BasicMatrix getACovariances(double[][] returns){
>>>
>>>   int row = returns.length;
>>>   int col = returns[0].length;
>>>   MatrixStore<BigDecimal> a = BigDenseStore.FACTORY..makeZero(row,
>>> col);
>>>   BasicMatrix covariances = new BigMatrix(a);
>>>
>>>   for(int i=1;i<=row;i++){
>>>    for(int j=i;j<=col;j++){
>>>     double tmp = this.getCovarance(returns[i-1], returns[j-1]);
>>>     covariances = covariances.set(i-1, j-1,tmp);
>>>     covariances = covariances.set(j-1, i-1,tmp);
>>>    }
>>>   }
>>>   return covariances;
>>>  }
>>>
>>>  public BasicMatrix getAnExpectedExcessReturns(double[][] returns){
>>>
>>>   int row = returns.length;
>>>   int col = returns[0].length;
>>>
>>>   MatrixStore<BigDecimal> a = BigDenseStore.FACTORY..makeZero 
>>> (row, 1);
>>>   BasicMatrix expectedExcessReturns = new BigMatrix(a);
>>>
>>>   double riskFreeReturn = 0;
>>>   for(int i=1;i<=col;i++)
>>>    riskFreeReturn+=returns[row-1][i-1];
>>>   riskFreeReturn/=row;
>>>
>>>   for(int i=1;i<=row;i++){
>>>
>>>    double tmp = 0;
>>>    for(int j=1;j<=col;j++)
>>>     tmp+=returns[i-1][j-1];
>>>    tmp/=col;
>>>
>>>    expectedExcessReturns = expectedExcessReturns.set(i-1, 0, tmp-
>>> riskFreeReturn);
>>>   }
>>>
>>>   return expectedExcessReturns;
>>>  }
>>>
>>>  public static void main(String[] args)
>>>  {
>>>   int assetNum=7;
>>>   double[][] assets_return={
>>>    {-1.5905837442343828E-4, -0.03062360801781757,
>>> -0.029857534032853142, -0.011811692726036832,
>>> -0.017972310602803136, 0.017338003502626997, 0.0},
>>>    {-0.02757158006362653, -0.02562704471101405,
>>> -0.011751538891997735, -0.024915062287655786, -0.01684088269454123,
>>> 0.013585351447135364, 0.0},
>>>    {-0.00699300699300693, -0.033802816901408676,
>>> -0.04675196850393671, -0.021166752710376546, -0.007911392405063583,
>>> 0.03827751196172254, 0.0},
>>>    {-0.007626310772164015, 0.0038424591738713027,
>>> 0.02488038277511978, 0.025210084033613675, -0.02003642987249557,
>>> -0.09758364312267642, 0.0},
>>>    {-0.03965053763440893, 0.021693491952414375,
>>> 0.01643835616438392, -0.007412398921833087, 0.01765105227427014,
>>> -0.010006671114076025, 0.0},
>>>    {-0.017821782178217872, 0.005040322580645311,
>>> 0.006018054162487363, 9.008107296569024E-4, 0.002999999999999824,
>>> -0.01196410767696908, 0.0},
>>>    {2.630552127527583E-4, 2.5867028174649627E-4,
>>> 2.3866431891514327E-4, 1.9564035993080523E-4, 2.351016690966669E-4,
>>> 1.9070675120065465E-4, 0.0}
>>>   };
>>>
>>>   TestMarkowitz tm = new TestMarkowitz();
>>>   BasicMatrix covariances = tm.getACovariances(assets_return);
>>>   BasicMatrix expectedExcessReturns = tm.getAnExpectedExcessReturns
>>> (assets_return);
>>>   BigDecimal riskAversion = new BigDecimal(1.0);
>>>
>>>   MarketEquilibrium marketEquilibrium = new MarketEquilibrium
>>> (covariances,riskAversion);
>>>   MarkowitzModel markowitzModel = new MarkowitzModel
>>> (marketEquilibrium,expectedExcessReturns);
>>>
>>>   for(int i=0;i<assetNum;i++){
>>>    markowitzModel.setLowerLimit(i, new BigDecimal(0.0));
>>>    markowitzModel.setUpperLimit(i, new BigDecimal(1.0));
>>>   }
>>>   List<BigDecimal> re = markowitzModel.getWeights();
>>>
>>>   System.out.println("=======result====================");
>>>   for(int i=0;i<re.size();i++)
>>>    System.out.println(re.get(i));
>>>
>>>   return;
>>>  }
>>>
>>> }
>>>
>>> /
>>> ******************************************************************** 
>>> **
>>> **********/
>>>
>>> and the result turn out as:
>>>
>>> -0.25
>>> -0.25
>>> -0.25
>>> -0.25
>>> 0.75
>>> 0.75
>>> 0.75
>>>
>>> /
>>> ******************************************************************** 
>>> **
>>> ******/
>>>
>>> If you find out any problem, just let me know , thanks a lot!
>>>
>>> On Thu, 15 Jan 2009 09:20:52 +0100, Anders Peterson wrote
>>>> You have to send code (snippets) that I can execute.
>>>>
>>>> Are you using v25 or the latest from cvs?
>>>>
>>>> How do you instantiate the ActiveSetSolver?
>>>>
>>>> What's the correct solution?
>>>>
>>>> You say that if you change the sign of "C" you get an ok result.
>>>> What  does this mean? Did you make a mistake with the sign, or do
>>>> you  suspect there is a bug in ojAlgo?
>>>>
>>>> I need better info regarding what you do, and what the problem is!
>>>>
>>>> /Anders
>>>>
>>>> On 15 jan 2009, at 06.14, Ruan Weichao wrote:
>>>>
>>>>>
>>>>> Thank you for your response.
>>>>>
>>>>> Here's my model's detail:
>>>>>
>>>>> [AE]={1.0,     1.0,     1.0,       1.0,     1.0,     1.0,
>>> 1.0}
>>>>>
>>>>>
>>>>> [be]={1.0}
>>>>>
>>>>>
>>>>> [Q]={3.048907897157133E-4, 1.6671472561019247E-4,
>>>>> 4.4500080981934345E-4, -5.389129745055723E-4,
>>>>> -2.6090705011393183E-4, -1.2633284900760366E-4,
>>>>> -6.485428846447651E-7},
>>>>> {1.6671472561019247E-4, 2.341985572849691E-4,
>>>>> 2.9113916450678265E-4, -4.5760873539850514E-4,
>>>>> 1.3078636134987255E-5, -2.354289901013046E-5,
>>> -7.578030042426654E-7},
>>>>> {4.4500080981934345E-4, 2.9113916450678265E-4,
>>> 7.46023915996829E-4,
>>>>> -0.0010247176498305568, -2.6745504327902895E-4,
>>>>> -1.6563544154823496E-4, -8.293698990696063E-7},
>>>>> {-5.389129745055723E-4, -4.5760873539850514E-4,
>>>>> -0.0010247176498305568, 0.001754169535149865,
>>>>> 2.0293065310212377E-4, 2.1401092557826588E-4,
>>> 1.0252846778608953E-7},
>>>>> {-2.6090705011393183E-4, 1.3078636134987255E-5,
>>>>> -2.6745504327902895E-4, 2.0293065310212377E-4,
>>>>> 4.632320892679136E-4, 1.7969731066037214E-4,
>>> 2.4953495129362833E-8},
>>>>> {-1.2633284900760366E-4, -2.354289901013046E-5,
>>>>> -1.6563544154823496E-4, 2.1401092557826588E-4,
>>>>> 1.7969731066037214E-4, 8..346410612364995E-5,
>>> -7.02099350897589E-8},
>>>>> {-6.485428846447651E-7, -7.578030042426654E-7,
>>>>> -8.293698990696063E-7, 1.0252846778608953E-7,
>>>>> 2.4953495129362833E-8, -7.02099350897589E-8,
>>> 8.367244992498656E-9}}
>>>>>
>>>>> [c]={{0.010638291263564232},
>>>>> {0.013500370827906071},
>>>>> {0.011390037735101773},
>>>>> {0.010385042339767682},
>>>>> {3.812208389845893E-4},
>>>>> {0.002315505853720011},
>>>>> {0.0}}
>>>>>
>>>>>
>>>>> [AI]={{1.0,     0.0,     0.0,       0.0,     0.0,
>>> 0.0,       0.0},
>>>>> {0.0,     1.0,     0.0,      0.0,     0.0,     0.0,       0.0},
>>>>> {0.0,     0.0,     1.0,      0.0,     0.0,     0.0,       0.0},
>>>>> {0.0,     0.0,     0.0,      1.0,     0.0,     0.0,       0.0},
>>>>> {0.0,     0.0,     0.0,      0.0,     1.0,     0.0,       0.0},
>>>>> {0.0,     0.0,     0.0,      0.0,     0.0,     1.0,       0.0},
>>>>> {0.0,     0.0,     0.0,      0.0,     0.0,     0.0,       1.0},
>>>>> {-1.0,     0.0,     0.0,      0.0,     0.0,     0.0,       0.0},
>>>>> {0.0,     -1.0,     0.0,      0.0,     0.0,     0.0,       0.0},
>>>>> {0.0,     0.0,     -1.0,      0.0,     0.0,     0.0,       0.0},
>>>>> {0.0,     0.0,     0.0,      -1.0,     0.0,     0.0,       0.0},
>>>>> {0.0,     0.0,     0.0,      0.0,     -1.0,     0.0,       0.0},
>>>>> {0.0,     0.0,     0.0,      0.0,     0.0,     -1.0,       0.0},
>>>>> {0.0,     0.0,     0.0,      0.0,     0.0,     0.0,       -1.0}}
>>>>>
>>>>> [bi]={{1.0},
>>>>> {1.0},
>>>>> {1.0},
>>>>> {1.0},
>>>>> {1.0},
>>>>> {1.0},
>>>>> {1.0},
>>>>> {0.0},
>>>>> {0.0},
>>>>> {0.0},
>>>>> {0.0},
>>>>> {0.0},
>>>>> {0.0},
>>>>> {0.0}}
>>>>>
>>>>> And here's the result:
>>>>>
>>>>> -0.25
>>>>> -0.25
>>>>> -0.25
>>>>> -0.25
>>>>> 0.75
>>>>> 0.75
>>>>> 0.75
>>>>>
>>>>> And i found that if i change the nagetive ExceptedReturn to
>>>>> positive, the result will be fine.
>>>>>
>>>>>
>>>>> On Wed, 14 Jan 2009 09:27:53 +0100, Anders Peterson wrote
>>>>>> How small are the negative values?
>>>>>>
>>>>>> Give me a unit test (something that I can easily integrate
>>> with the
>>>>>> existing junit tests) that shows what you do, and where/how you
>>>>>> think  it goes wrong.
>>>>>>
>>>>>> /Anders
>>>>>>
>>>>>> On 14 jan 2009, at 03.58, Ruan Weichao wrote:
>>>>>>
>>>>>>>
>>>>>>> Hi,
>>>>>>>
>>>>>>> I use the  Markowitz model to calculate the optimize some
>>> assets'
>>>>>>> weights,
>>>>>>>
>>>>>>> and set all the assets' lowerlimit with Zero and set all the
>>>>>>> upperlimits with
>>>>>>>
>>>>>>> One, but it turns out with some negative value which are
>>> little
>>>>>>> than Zero, it
>>>>>>>
>>>>>>> seems that the lowerlimit does not work.
>>>>>>>
>>>>>>> So I want to ask what should I do it to confirm the
>>> lowerlimit?
>>>>>>>
>>>>>>> --
>>>>>>> Best regards
>>>>>>>
>>>>>>> weichao
>>>>>>>
>>>>>>>
>>>>>
>>> -------------------------------------------------------------------- 
>>> --
>>>>>>> --------
>>>>>>> This SF.net email is sponsored by:
>>>>>>> SourcForge Community
>>>>>>> SourceForge wants to tell your story.
>>>>>>> http://p.sf.net/sfu/sf-spreadtheword
>>>>>>> _______________________________________________
>>>>>>> ojAlgo-user mailing list
>>>>>>> ojA...@li...
>>>>>>> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>
>>> -------------------------------------------------------------------- 
>>> --
>>>>> ------
>>>>> --
>>>>>> This SF.net email is sponsored by:
>>>>>> SourcForge Community
>>>>>> SourceForge wants to tell your story.
>>>>>> http://p.sf.net/sfu/sf-spreadtheword
>>>>>> _______________________________________________
>>>>>> ojAlgo-user mailing list
>>>>>> ojA...@li...
>>>>>> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
>>>>>
>>>>>
>>>>> --
>>>>> Best regards
>>>>>
>>>>>
>>> -------------------------------------------------------------------- 
>>> --
>>>>> --------
>>>>> This SF.net email is sponsored by:
>>>>> SourcForge Community
>>>>> SourceForge wants to tell your story.
>>>>> http://p.sf.net/sfu/sf-
>>>>> spreadtheword_______________________________________________
>>>>> ojAlgo-user mailing list
>>>>> ojA...@li...
>>>>> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
>>>>
>>>>
>>> -------------------------------------------------------------------- 
>>> --
>>> ------
>>> -- 
>>>> This SF.net email is sponsored by:
>>>> SourcForge Community
>>>> SourceForge wants to tell your story.
>>>> http://p.sf.net/sfu/sf-spreadtheword
>>>> _______________________________________________
>>>> ojAlgo-user mailing list
>>>> ojA...@li...
>>>> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
>>>
>>>
>>> -- 
>>> Best regards
>>>
>>> -------------------------------------------------------------------- 
>>> --
>>> --------
>>> This SF.net email is sponsored by:
>>> SourcForge Community
>>> SourceForge wants to tell your story.
>>> http://p.sf.net/sfu/sf-
>>> spreadtheword_______________________________________________
>>> ojAlgo-user mailing list
>>> ojA...@li...
>>> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
>>
>> --------------------------------------------------------------------- 
>> -------
> --
>> This SF.net email is sponsored by:
>> SourcForge Community
>> SourceForge wants to tell your story.
>> http://p.sf.net/sfu/sf-spreadtheword
>> _______________________________________________
>> ojAlgo-user mailing list
>> ojA...@li...
>> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
>
>
> --
> Best regards
>
>
> ---------------------------------------------------------------------- 
> --------
> This SF.net email is sponsored by:
> SourcForge Community
> SourceForge wants to tell your story.
> http://p.sf.net/sfu/sf-spreadtheword
> _______________________________________________
> ojAlgo-user mailing list
> ojA...@li...
> https://lists.sourceforge.net/lists/listinfo/ojalgo-user

Re: [ojAlgo-user] Markowitz Model Problem

[Ruan W. <is0...@ma...>]

I think i get your point.

So I should not use negative excess returns in my model.

Thank you very much!

On Fri, 16 Jan 2009 07:25:41 +0100, Anders Peterson wrote
> On 16 jan 2009, at 02.28, Ruan Weichao wrote:
> 
> > Hi,
> >
> > Yes, that's the expected excess returns of my program.
> >
> > Is it right?
> >
> > Can the "expected excess returns" be negative?
> 
> I have no way of knowing what model/problem you want to solve...
> 
> In a general case excess returns (what you get for taking a risk)  
> should not be negative. If they are, you do not need to optimise  
> anything. Just invest everything in the risk free asset.
> 
> I'm not sure exactly why you get those negative (constraint breaking)
>   portfolio weights. I'm going to look into this, but later. Don't 
> hold  your breath. For you, I think the solution is to work on your model.
> 
> /Anders
> 
> > On Thu, 15 Jan 2009 14:09:43 +0100, Anders Peterson wrote
> >> Your expected excess returns matrix looks like this:
> >>
> >> class org.ojalgo.matrix.BigMatrix
> >> {{-0.010638291263564232},
> >> {-0.013500370827906071},
> >> {-0.011390037735101773},
> >> {-0.010385042339767682},
> >> {-0.0003812208389845893},
> >> {-0.002315505853720011},
> >> {0}}
> >>
> >> Is that correct?
> >>
> >> Earlier you wrote "And i found that if i change the nagetive
> >> ExceptedReturn to positive, the result will be fine."
> >>
> >> /Anders
> >>
> >> On 15 jan 2009, at 11.10, Ruan Weichao wrote:
> >>
> >>> Hi，i am using v25 and here's my code:
> >>>
> >>> /
> >>> ******************************************************************** 
> >>> **
> >>> **********/
> >>>
> >>> import java.math.BigDecimal;
> >>> import java.util.List;
> >>>
> >>> import org.ojalgo.finance.portfolio.MarketEquilibrium;
> >>> import org.ojalgo.finance.portfolio.MarkowitzModel;
> >>> import org.ojalgo.matrix.*;
> >>> import org.ojalgo.matrix.store.BigDenseStore;
> >>> import org.ojalgo.matrix.store.MatrixStore;
> >>>
> >>>
> >>> public class TestMarkowitz {
> >>>
> >>>  public double getCovarance(double[] value1, double[] value2){
> >>>
> >>>   int n = value1.length;
> >>>   double averOne = 0;
> >>>   double averTwo = 0;
> >>>   for(int i=0;i<n;i++)
> >>>   {
> >>>    averOne+=value1[i];
> >>>    averTwo+=value2[i];
> >>>   }
> >>>   averOne /= n;
> >>>   averTwo /= n;
> >>>
> >>>   double sum = 0;
> >>>   for(int i=0;i<n;i++)
> >>>    sum+=(value1[i]-averOne)*(value2[i]-averTwo);
> >>>
> >>>   return sum / (n-1);
> >>>  }
> >>>
> >>>  public BasicMatrix getACovariances(double[][] returns){
> >>>
> >>>   int row = returns.length;
> >>>   int col = returns[0].length;
> >>>   MatrixStore<BigDecimal> a = BigDenseStore.FACTORY..makeZero(row,
> >>> col);
> >>>   BasicMatrix covariances = new BigMatrix(a);
> >>>
> >>>   for(int i=1;i<=row;i++){
> >>>    for(int j=i;j<=col;j++){
> >>>     double tmp = this.getCovarance(returns[i-1], returns[j-1]);
> >>>     covariances = covariances.set(i-1, j-1,tmp);
> >>>     covariances = covariances.set(j-1, i-1,tmp);
> >>>    }
> >>>   }
> >>>   return covariances;
> >>>  }
> >>>
> >>>  public BasicMatrix getAnExpectedExcessReturns(double[][] returns){
> >>>
> >>>   int row = returns.length;
> >>>   int col = returns[0].length;
> >>>
> >>>   MatrixStore<BigDecimal> a = BigDenseStore.FACTORY..makeZero 
> >>> (row, 1);
> >>>   BasicMatrix expectedExcessReturns = new BigMatrix(a);
> >>>
> >>>   double riskFreeReturn = 0;
> >>>   for(int i=1;i<=col;i++)
> >>>    riskFreeReturn+=returns[row-1][i-1];
> >>>   riskFreeReturn/=row;
> >>>
> >>>   for(int i=1;i<=row;i++){
> >>>
> >>>    double tmp = 0;
> >>>    for(int j=1;j<=col;j++)
> >>>     tmp+=returns[i-1][j-1];
> >>>    tmp/=col;
> >>>
> >>>    expectedExcessReturns = expectedExcessReturns.set(i-1, 0, tmp-
> >>> riskFreeReturn);
> >>>   }
> >>>
> >>>   return expectedExcessReturns;
> >>>  }
> >>>
> >>>  public static void main(String[] args)
> >>>  {
> >>>   int assetNum=7;
> >>>   double[][] assets_return={
> >>>    {-1.5905837442343828E-4, -0.03062360801781757,
> >>> -0.029857534032853142, -0.011811692726036832,
> >>> -0.017972310602803136, 0.017338003502626997, 0.0},
> >>>    {-0.02757158006362653, -0.02562704471101405,
> >>> -0.011751538891997735, -0.024915062287655786, -0.01684088269454123,
> >>> 0.013585351447135364, 0.0},
> >>>    {-0.00699300699300693, -0.033802816901408676,
> >>> -0.04675196850393671, -0.021166752710376546, -0.007911392405063583,
> >>> 0.03827751196172254, 0.0},
> >>>    {-0.007626310772164015, 0.0038424591738713027,
> >>> 0.02488038277511978, 0.025210084033613675, -0.02003642987249557,
> >>> -0.09758364312267642, 0.0},
> >>>    {-0.03965053763440893, 0.021693491952414375,
> >>> 0.01643835616438392, -0.007412398921833087, 0.01765105227427014,
> >>> -0.010006671114076025, 0.0},
> >>>    {-0.017821782178217872, 0.005040322580645311,
> >>> 0.006018054162487363, 9.008107296569024E-4, 0.002999999999999824,
> >>> -0.01196410767696908, 0.0},
> >>>    {2.630552127527583E-4, 2.5867028174649627E-4,
> >>> 2.3866431891514327E-4, 1.9564035993080523E-4, 2.351016690966669E-4,
> >>> 1.9070675120065465E-4, 0.0}
> >>>   };
> >>>
> >>>   TestMarkowitz tm = new TestMarkowitz();
> >>>   BasicMatrix covariances = tm.getACovariances(assets_return);
> >>>   BasicMatrix expectedExcessReturns = tm.getAnExpectedExcessReturns
> >>> (assets_return);
> >>>   BigDecimal riskAversion = new BigDecimal(1.0);
> >>>
> >>>   MarketEquilibrium marketEquilibrium = new MarketEquilibrium
> >>> (covariances,riskAversion);
> >>>   MarkowitzModel markowitzModel = new MarkowitzModel
> >>> (marketEquilibrium,expectedExcessReturns);
> >>>
> >>>   for(int i=0;i<assetNum;i++){
> >>>    markowitzModel.setLowerLimit(i, new BigDecimal(0.0));
> >>>    markowitzModel.setUpperLimit(i, new BigDecimal(1.0));
> >>>   }
> >>>   List<BigDecimal> re = markowitzModel.getWeights();
> >>>
> >>>   System.out.println("=======result====================");
> >>>   for(int i=0;i<re.size();i++)
> >>>    System.out.println(re.get(i));
> >>>
> >>>   return;
> >>>  }
> >>>
> >>> }
> >>>
> >>> /
> >>> ******************************************************************** 
> >>> **
> >>> **********/
> >>>
> >>> and the result turn out as:
> >>>
> >>> -0.25
> >>> -0.25
> >>> -0.25
> >>> -0.25
> >>> 0.75
> >>> 0.75
> >>> 0.75
> >>>
> >>> /
> >>> ******************************************************************** 
> >>> **
> >>> ******/
> >>>
> >>> If you find out any problem, just let me know , thanks a lot!
> >>>
> >>> On Thu, 15 Jan 2009 09:20:52 +0100, Anders Peterson wrote
> >>>> You have to send code (snippets) that I can execute.
> >>>>
> >>>> Are you using v25 or the latest from cvs?
> >>>>
> >>>> How do you instantiate the ActiveSetSolver?
> >>>>
> >>>> What's the correct solution?
> >>>>
> >>>> You say that if you change the sign of "C" you get an ok result.
> >>>> What  does this mean? Did you make a mistake with the sign, or do
> >>>> you  suspect there is a bug in ojAlgo?
> >>>>
> >>>> I need better info regarding what you do, and what the problem is!
> >>>>
> >>>> /Anders
> >>>>
> >>>> On 15 jan 2009, at 06.14, Ruan Weichao wrote:
> >>>>
> >>>>>
> >>>>> Thank you for your response.
> >>>>>
> >>>>> Here's my model's detail:
> >>>>>
> >>>>> [AE]={1.0,     1.0,     1.0,       1.0,     1.0,     1.0,
> >>> 1.0}
> >>>>>
> >>>>>
> >>>>> [be]={1.0}
> >>>>>
> >>>>>
> >>>>> [Q]={3.048907897157133E-4, 1.6671472561019247E-4,
> >>>>> 4.4500080981934345E-4, -5.389129745055723E-4,
> >>>>> -2.6090705011393183E-4, -1.2633284900760366E-4,
> >>>>> -6.485428846447651E-7},
> >>>>> {1.6671472561019247E-4, 2.341985572849691E-4,
> >>>>> 2.9113916450678265E-4, -4.5760873539850514E-4,
> >>>>> 1.3078636134987255E-5, -2.354289901013046E-5,
> >>> -7.578030042426654E-7},
> >>>>> {4.4500080981934345E-4, 2.9113916450678265E-4,
> >>> 7.46023915996829E-4,
> >>>>> -0.0010247176498305568, -2.6745504327902895E-4,
> >>>>> -1.6563544154823496E-4, -8.293698990696063E-7},
> >>>>> {-5.389129745055723E-4, -4.5760873539850514E-4,
> >>>>> -0.0010247176498305568, 0.001754169535149865,
> >>>>> 2.0293065310212377E-4, 2.1401092557826588E-4,
> >>> 1.0252846778608953E-7},
> >>>>> {-2.6090705011393183E-4, 1.3078636134987255E-5,
> >>>>> -2.6745504327902895E-4, 2.0293065310212377E-4,
> >>>>> 4.632320892679136E-4, 1.7969731066037214E-4,
> >>> 2.4953495129362833E-8},
> >>>>> {-1.2633284900760366E-4, -2.354289901013046E-5,
> >>>>> -1.6563544154823496E-4, 2.1401092557826588E-4,
> >>>>> 1.7969731066037214E-4, 8..346410612364995E-5,
> >>> -7.02099350897589E-8},
> >>>>> {-6.485428846447651E-7, -7.578030042426654E-7,
> >>>>> -8.293698990696063E-7, 1.0252846778608953E-7,
> >>>>> 2.4953495129362833E-8, -7.02099350897589E-8,
> >>> 8.367244992498656E-9}}
> >>>>>
> >>>>> [c]={{0.010638291263564232},
> >>>>> {0.013500370827906071},
> >>>>> {0.011390037735101773},
> >>>>> {0.010385042339767682},
> >>>>> {3.812208389845893E-4},
> >>>>> {0.002315505853720011},
> >>>>> {0.0}}
> >>>>>
> >>>>>
> >>>>> [AI]={{1.0,     0.0,     0.0,       0.0,     0.0,
> >>> 0.0,       0.0},
> >>>>> {0.0,     1.0,     0.0,      0.0,     0.0,     0.0,       0.0},
> >>>>> {0.0,     0.0,     1.0,      0.0,     0.0,     0.0,       0.0},
> >>>>> {0.0,     0.0,     0.0,      1.0,     0.0,     0.0,       0.0},
> >>>>> {0.0,     0.0,     0.0,      0.0,     1.0,     0.0,       0.0},
> >>>>> {0.0,     0.0,     0.0,      0.0,     0.0,     1.0,       0.0},
> >>>>> {0.0,     0.0,     0.0,      0.0,     0.0,     0.0,       1.0},
> >>>>> {-1.0,     0.0,     0.0,      0.0,     0.0,     0.0,       0.0},
> >>>>> {0.0,     -1.0,     0.0,      0.0,     0.0,     0.0,       0.0},
> >>>>> {0.0,     0.0,     -1.0,      0.0,     0.0,     0.0,       0.0},
> >>>>> {0.0,     0.0,     0.0,      -1.0,     0.0,     0.0,       0.0},
> >>>>> {0.0,     0.0,     0.0,      0.0,     -1.0,     0.0,       0.0},
> >>>>> {0.0,     0.0,     0.0,      0.0,     0.0,     -1.0,       0.0},
> >>>>> {0.0,     0.0,     0.0,      0.0,     0.0,     0.0,       -1.0}}
> >>>>>
> >>>>> [bi]={{1.0},
> >>>>> {1.0},
> >>>>> {1.0},
> >>>>> {1.0},
> >>>>> {1.0},
> >>>>> {1.0},
> >>>>> {1.0},
> >>>>> {0.0},
> >>>>> {0.0},
> >>>>> {0.0},
> >>>>> {0.0},
> >>>>> {0.0},
> >>>>> {0.0},
> >>>>> {0.0}}
> >>>>>
> >>>>> And here's the result:
> >>>>>
> >>>>> -0.25
> >>>>> -0.25
> >>>>> -0.25
> >>>>> -0.25
> >>>>> 0.75
> >>>>> 0.75
> >>>>> 0.75
> >>>>>
> >>>>> And i found that if i change the nagetive ExceptedReturn to
> >>>>> positive, the result will be fine.
> >>>>>
> >>>>>
> >>>>> On Wed, 14 Jan 2009 09:27:53 +0100, Anders Peterson wrote
> >>>>>> How small are the negative values?
> >>>>>>
> >>>>>> Give me a unit test (something that I can easily integrate
> >>> with the
> >>>>>> existing junit tests) that shows what you do, and where/how you
> >>>>>> think  it goes wrong.
> >>>>>>
> >>>>>> /Anders
> >>>>>>
> >>>>>> On 14 jan 2009, at 03.58, Ruan Weichao wrote:
> >>>>>>
> >>>>>>>
> >>>>>>> Hi,
> >>>>>>>
> >>>>>>> I use the  Markowitz model to calculate the optimize some
> >>> assets'
> >>>>>>> weights,
> >>>>>>>
> >>>>>>> and set all the assets' lowerlimit with Zero and set all the
> >>>>>>> upperlimits with
> >>>>>>>
> >>>>>>> One, but it turns out with some negative value which are
> >>> little
> >>>>>>> than Zero, it
> >>>>>>>
> >>>>>>> seems that the lowerlimit does not work.
> >>>>>>>
> >>>>>>> So I want to ask what should I do it to confirm the
> >>> lowerlimit?
> >>>>>>>
> >>>>>>> --
> >>>>>>> Best regards
> >>>>>>>
> >>>>>>> weichao
> >>>>>>>
> >>>>>>>
> >>>>>
> >>> -------------------------------------------------------------------- 
> >>> --
> >>>>>>> --------
> >>>>>>> This SF.net email is sponsored by:
> >>>>>>> SourcForge Community
> >>>>>>> SourceForge wants to tell your story.
> >>>>>>> http://p.sf.net/sfu/sf-spreadtheword
> >>>>>>> _______________________________________________
> >>>>>>> ojAlgo-user mailing list
> >>>>>>> ojA...@li...
> >>>>>>> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
> >>>>>>>
> >>>>>>>
> >>>>>>
> >>>>>>
> >>>>>
> >>> -------------------------------------------------------------------- 
> >>> --
> >>>>> ------
> >>>>> --
> >>>>>> This SF.net email is sponsored by:
> >>>>>> SourcForge Community
> >>>>>> SourceForge wants to tell your story.
> >>>>>> http://p.sf.net/sfu/sf-spreadtheword
> >>>>>> _______________________________________________
> >>>>>> ojAlgo-user mailing list
> >>>>>> ojA...@li...
> >>>>>> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
> >>>>>
> >>>>>
> >>>>> --
> >>>>> Best regards
> >>>>>
> >>>>>
> >>> -------------------------------------------------------------------- 
> >>> --
> >>>>> --------
> >>>>> This SF.net email is sponsored by:
> >>>>> SourcForge Community
> >>>>> SourceForge wants to tell your story.
> >>>>> http://p.sf.net/sfu/sf-
> >>>>> spreadtheword_______________________________________________
> >>>>> ojAlgo-user mailing list
> >>>>> ojA...@li...
> >>>>> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
> >>>>
> >>>>
> >>> -------------------------------------------------------------------- 
> >>> --
> >>> ------
> >>> -- 
> >>>> This SF.net email is sponsored by:
> >>>> SourcForge Community
> >>>> SourceForge wants to tell your story.
> >>>> http://p.sf.net/sfu/sf-spreadtheword
> >>>> _______________________________________________
> >>>> ojAlgo-user mailing list
> >>>> ojA...@li...
> >>>> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
> >>>
> >>>
> >>> -- 
> >>> Best regards
> >>>
> >>> -------------------------------------------------------------------- 
> >>> --
> >>> --------
> >>> This SF.net email is sponsored by:
> >>> SourcForge Community
> >>> SourceForge wants to tell your story.
> >>> http://p.sf.net/sfu/sf-
> >>> spreadtheword_______________________________________________
> >>> ojAlgo-user mailing list
> >>> ojA...@li...
> >>> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
> >>
> >> --------------------------------------------------------------------- 
> >> -------
> > --
> >> This SF.net email is sponsored by:
> >> SourcForge Community
> >> SourceForge wants to tell your story.
> >> http://p.sf.net/sfu/sf-spreadtheword
> >> _______________________________________________
> >> ojAlgo-user mailing list
> >> ojA...@li...
> >> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
> >
> >
> > --
> > Best regards
> >
> >
> > ---------------------------------------------------------------------- 
> > --------
> > This SF.net email is sponsored by:
> > SourcForge Community
> > SourceForge wants to tell your story.
> > http://p.sf.net/sfu/sf-spreadtheword
> > _______________________________________________
> > ojAlgo-user mailing list
> > ojA...@li...
> > https://lists.sourceforge.net/lists/listinfo/ojalgo-user
> 
> ----------------------------------------------------------------------------
--
> This SF.net email is sponsored by:
> SourcForge Community
> SourceForge wants to tell your story.
> http://p.sf.net/sfu/sf-spreadtheword
> _______________________________________________
> ojAlgo-user mailing list
> ojA...@li...
> https://lists.sourceforge.net/lists/listinfo/ojalgo-user


--
Best regards

Re: [ojAlgo-user] Markowitz Model Problem

[Anders P. <an...@op...>]

On 16 jan 2009, at 07.25, Anders Peterson wrote:

> On 16 jan 2009, at 02.28, Ruan Weichao wrote:
>
>> Hi,
>>
>> Yes, that's the expected excess returns of my program.
>>
>> Is it right?
>>
>> Can the "expected excess returns" be negative?
>
>
> I have no way of knowing what model/problem you want to solve...
>
> In a general case excess returns (what you get for taking a risk)  
> should not be negative. If they are, you do not need to optimise  
> anything. Just invest everything in the risk free asset.
>
> I'm not sure exactly why you get those negative (constraint  
> breaking) portfolio weights. I'm going to look into this, but  
> later. Don't hold your breath. For you, I think the solution is to  
> work on your model.


This is fixed in CVS now. Your model does not get those constraint  
breaking portfolio weights anymore. The solution I get is 100% of the  
last instrument (the one with 0.0 excess return) which seems reasonable.

/Anders


>
>> On Thu, 15 Jan 2009 14:09:43 +0100, Anders Peterson wrote
>>> Your expected excess returns matrix looks like this:
>>>
>>> class org.ojalgo.matrix.BigMatrix
>>> {{-0.010638291263564232},
>>> {-0.013500370827906071},
>>> {-0.011390037735101773},
>>> {-0.010385042339767682},
>>> {-0.0003812208389845893},
>>> {-0.002315505853720011},
>>> {0}}
>>>
>>> Is that correct?
>>>
>>> Earlier you wrote "And i found that if i change the nagetive
>>> ExceptedReturn to positive, the result will be fine."
>>>
>>> /Anders
>>>
>>> On 15 jan 2009, at 11.10, Ruan Weichao wrote:
>>>
>>>> Hi，i am using v25 and here's my code:
>>>>
>>>> /
>>>> ******************************************************************* 
>>>> ***
>>>> **********/
>>>>
>>>> import java.math.BigDecimal;
>>>> import java.util.List;
>>>>
>>>> import org.ojalgo.finance.portfolio.MarketEquilibrium;
>>>> import org.ojalgo.finance.portfolio.MarkowitzModel;
>>>> import org.ojalgo.matrix.*;
>>>> import org.ojalgo.matrix.store.BigDenseStore;
>>>> import org.ojalgo.matrix.store.MatrixStore;
>>>>
>>>>
>>>> public class TestMarkowitz {
>>>>
>>>>  public double getCovarance(double[] value1, double[] value2){
>>>>
>>>>   int n = value1.length;
>>>>   double averOne = 0;
>>>>   double averTwo = 0;
>>>>   for(int i=0;i<n;i++)
>>>>   {
>>>>    averOne+=value1[i];
>>>>    averTwo+=value2[i];
>>>>   }
>>>>   averOne /= n;
>>>>   averTwo /= n;
>>>>
>>>>   double sum = 0;
>>>>   for(int i=0;i<n;i++)
>>>>    sum+=(value1[i]-averOne)*(value2[i]-averTwo);
>>>>
>>>>   return sum / (n-1);
>>>>  }
>>>>
>>>>  public BasicMatrix getACovariances(double[][] returns){
>>>>
>>>>   int row = returns.length;
>>>>   int col = returns[0].length;
>>>>   MatrixStore<BigDecimal> a = BigDenseStore.FACTORY..makeZero(row,
>>>> col);
>>>>   BasicMatrix covariances = new BigMatrix(a);
>>>>
>>>>   for(int i=1;i<=row;i++){
>>>>    for(int j=i;j<=col;j++){
>>>>     double tmp = this.getCovarance(returns[i-1], returns[j-1]);
>>>>     covariances = covariances.set(i-1, j-1,tmp);
>>>>     covariances = covariances.set(j-1, i-1,tmp);
>>>>    }
>>>>   }
>>>>   return covariances;
>>>>  }
>>>>
>>>>  public BasicMatrix getAnExpectedExcessReturns(double[][] returns){
>>>>
>>>>   int row = returns.length;
>>>>   int col = returns[0].length;
>>>>
>>>>   MatrixStore<BigDecimal> a = BigDenseStore.FACTORY..makeZero 
>>>> (row, 1);
>>>>   BasicMatrix expectedExcessReturns = new BigMatrix(a);
>>>>
>>>>   double riskFreeReturn = 0;
>>>>   for(int i=1;i<=col;i++)
>>>>    riskFreeReturn+=returns[row-1][i-1];
>>>>   riskFreeReturn/=row;
>>>>
>>>>   for(int i=1;i<=row;i++){
>>>>
>>>>    double tmp = 0;
>>>>    for(int j=1;j<=col;j++)
>>>>     tmp+=returns[i-1][j-1];
>>>>    tmp/=col;
>>>>
>>>>    expectedExcessReturns = expectedExcessReturns.set(i-1, 0, tmp-
>>>> riskFreeReturn);
>>>>   }
>>>>
>>>>   return expectedExcessReturns;
>>>>  }
>>>>
>>>>  public static void main(String[] args)
>>>>  {
>>>>   int assetNum=7;
>>>>   double[][] assets_return={
>>>>    {-1.5905837442343828E-4, -0.03062360801781757,
>>>> -0.029857534032853142, -0.011811692726036832,
>>>> -0.017972310602803136, 0.017338003502626997, 0.0},
>>>>    {-0.02757158006362653, -0.02562704471101405,
>>>> -0.011751538891997735, -0.024915062287655786, -0.01684088269454123,
>>>> 0.013585351447135364, 0.0},
>>>>    {-0.00699300699300693, -0.033802816901408676,
>>>> -0.04675196850393671, -0.021166752710376546, -0.007911392405063583,
>>>> 0.03827751196172254, 0.0},
>>>>    {-0.007626310772164015, 0.0038424591738713027,
>>>> 0.02488038277511978, 0.025210084033613675, -0.02003642987249557,
>>>> -0.09758364312267642, 0.0},
>>>>    {-0.03965053763440893, 0.021693491952414375,
>>>> 0.01643835616438392, -0.007412398921833087, 0.01765105227427014,
>>>> -0.010006671114076025, 0.0},
>>>>    {-0.017821782178217872, 0.005040322580645311,
>>>> 0.006018054162487363, 9.008107296569024E-4, 0.002999999999999824,
>>>> -0.01196410767696908, 0.0},
>>>>    {2.630552127527583E-4, 2.5867028174649627E-4,
>>>> 2.3866431891514327E-4, 1.9564035993080523E-4, 2.351016690966669E-4,
>>>> 1.9070675120065465E-4, 0.0}
>>>>   };
>>>>
>>>>   TestMarkowitz tm = new TestMarkowitz();
>>>>   BasicMatrix covariances = tm.getACovariances(assets_return);
>>>>   BasicMatrix expectedExcessReturns = tm.getAnExpectedExcessReturns
>>>> (assets_return);
>>>>   BigDecimal riskAversion = new BigDecimal(1.0);
>>>>
>>>>   MarketEquilibrium marketEquilibrium = new MarketEquilibrium
>>>> (covariances,riskAversion);
>>>>   MarkowitzModel markowitzModel = new MarkowitzModel
>>>> (marketEquilibrium,expectedExcessReturns);
>>>>
>>>>   for(int i=0;i<assetNum;i++){
>>>>    markowitzModel.setLowerLimit(i, new BigDecimal(0.0));
>>>>    markowitzModel.setUpperLimit(i, new BigDecimal(1.0));
>>>>   }
>>>>   List<BigDecimal> re = markowitzModel.getWeights();
>>>>
>>>>   System.out.println("=======result====================");
>>>>   for(int i=0;i<re.size();i++)
>>>>    System.out.println(re.get(i));
>>>>
>>>>   return;
>>>>  }
>>>>
>>>> }
>>>>
>>>> /
>>>> ******************************************************************* 
>>>> ***
>>>> **********/
>>>>
>>>> and the result turn out as:
>>>>
>>>> -0.25
>>>> -0.25
>>>> -0.25
>>>> -0.25
>>>> 0.75
>>>> 0.75
>>>> 0.75
>>>>
>>>> /
>>>> ******************************************************************* 
>>>> ***
>>>> ******/
>>>>
>>>> If you find out any problem, just let me know , thanks a lot!
>>>>
>>>> On Thu, 15 Jan 2009 09:20:52 +0100, Anders Peterson wrote
>>>>> You have to send code (snippets) that I can execute.
>>>>>
>>>>> Are you using v25 or the latest from cvs?
>>>>>
>>>>> How do you instantiate the ActiveSetSolver?
>>>>>
>>>>> What's the correct solution?
>>>>>
>>>>> You say that if you change the sign of "C" you get an ok result.
>>>>> What  does this mean? Did you make a mistake with the sign, or do
>>>>> you  suspect there is a bug in ojAlgo?
>>>>>
>>>>> I need better info regarding what you do, and what the problem is!
>>>>>
>>>>> /Anders
>>>>>
>>>>> On 15 jan 2009, at 06.14, Ruan Weichao wrote:
>>>>>
>>>>>>
>>>>>> Thank you for your response.
>>>>>>
>>>>>> Here's my model's detail:
>>>>>>
>>>>>> [AE]={1.0,     1.0,     1.0,       1.0,     1.0,     1.0,
>>>> 1.0}
>>>>>>
>>>>>>
>>>>>> [be]={1.0}
>>>>>>
>>>>>>
>>>>>> [Q]={3.048907897157133E-4, 1.6671472561019247E-4,
>>>>>> 4.4500080981934345E-4, -5.389129745055723E-4,
>>>>>> -2.6090705011393183E-4, -1.2633284900760366E-4,
>>>>>> -6.485428846447651E-7},
>>>>>> {1.6671472561019247E-4, 2.341985572849691E-4,
>>>>>> 2.9113916450678265E-4, -4.5760873539850514E-4,
>>>>>> 1.3078636134987255E-5, -2.354289901013046E-5,
>>>> -7.578030042426654E-7},
>>>>>> {4.4500080981934345E-4, 2.9113916450678265E-4,
>>>> 7.46023915996829E-4,
>>>>>> -0.0010247176498305568, -2.6745504327902895E-4,
>>>>>> -1.6563544154823496E-4, -8.293698990696063E-7},
>>>>>> {-5.389129745055723E-4, -4.5760873539850514E-4,
>>>>>> -0.0010247176498305568, 0.001754169535149865,
>>>>>> 2.0293065310212377E-4, 2.1401092557826588E-4,
>>>> 1.0252846778608953E-7},
>>>>>> {-2.6090705011393183E-4, 1.3078636134987255E-5,
>>>>>> -2.6745504327902895E-4, 2.0293065310212377E-4,
>>>>>> 4.632320892679136E-4, 1.7969731066037214E-4,
>>>> 2.4953495129362833E-8},
>>>>>> {-1.2633284900760366E-4, -2.354289901013046E-5,
>>>>>> -1.6563544154823496E-4, 2.1401092557826588E-4,
>>>>>> 1.7969731066037214E-4, 8..346410612364995E-5,
>>>> -7.02099350897589E-8},
>>>>>> {-6.485428846447651E-7, -7.578030042426654E-7,
>>>>>> -8.293698990696063E-7, 1.0252846778608953E-7,
>>>>>> 2.4953495129362833E-8, -7.02099350897589E-8,
>>>> 8.367244992498656E-9}}
>>>>>>
>>>>>> [c]={{0.010638291263564232},
>>>>>> {0.013500370827906071},
>>>>>> {0.011390037735101773},
>>>>>> {0.010385042339767682},
>>>>>> {3.812208389845893E-4},
>>>>>> {0.002315505853720011},
>>>>>> {0.0}}
>>>>>>
>>>>>>
>>>>>> [AI]={{1.0,     0.0,     0.0,       0.0,     0.0,
>>>> 0.0,       0.0},
>>>>>> {0.0,     1.0,     0.0,      0.0,     0.0,     0.0,       0.0},
>>>>>> {0.0,     0.0,     1.0,      0.0,     0.0,     0.0,       0.0},
>>>>>> {0.0,     0.0,     0.0,      1.0,     0.0,     0.0,       0.0},
>>>>>> {0.0,     0.0,     0.0,      0.0,     1.0,     0.0,       0.0},
>>>>>> {0.0,     0.0,     0.0,      0.0,     0.0,     1.0,       0.0},
>>>>>> {0.0,     0.0,     0.0,      0.0,     0.0,     0.0,       1.0},
>>>>>> {-1.0,     0.0,     0.0,      0.0,     0.0,     0.0,       0.0},
>>>>>> {0.0,     -1.0,     0.0,      0.0,     0.0,     0.0,       0.0},
>>>>>> {0.0,     0.0,     -1.0,      0.0,     0.0,     0.0,       0.0},
>>>>>> {0.0,     0.0,     0.0,      -1.0,     0.0,     0.0,       0.0},
>>>>>> {0.0,     0.0,     0.0,      0.0,     -1.0,     0.0,       0.0},
>>>>>> {0.0,     0.0,     0.0,      0.0,     0.0,     -1.0,       0.0},
>>>>>> {0.0,     0.0,     0.0,      0.0,     0.0,     0.0,       -1.0}}
>>>>>>
>>>>>> [bi]={{1.0},
>>>>>> {1.0},
>>>>>> {1.0},
>>>>>> {1.0},
>>>>>> {1.0},
>>>>>> {1.0},
>>>>>> {1.0},
>>>>>> {0.0},
>>>>>> {0.0},
>>>>>> {0.0},
>>>>>> {0.0},
>>>>>> {0.0},
>>>>>> {0.0},
>>>>>> {0.0}}
>>>>>>
>>>>>> And here's the result:
>>>>>>
>>>>>> -0.25
>>>>>> -0.25
>>>>>> -0.25
>>>>>> -0.25
>>>>>> 0.75
>>>>>> 0.75
>>>>>> 0.75
>>>>>>
>>>>>> And i found that if i change the nagetive ExceptedReturn to
>>>>>> positive, the result will be fine.
>>>>>>
>>>>>>
>>>>>> On Wed, 14 Jan 2009 09:27:53 +0100, Anders Peterson wrote
>>>>>>> How small are the negative values?
>>>>>>>
>>>>>>> Give me a unit test (something that I can easily integrate
>>>> with the
>>>>>>> existing junit tests) that shows what you do, and where/how you
>>>>>>> think  it goes wrong.
>>>>>>>
>>>>>>> /Anders
>>>>>>>
>>>>>>> On 14 jan 2009, at 03.58, Ruan Weichao wrote:
>>>>>>>
>>>>>>>>
>>>>>>>> Hi,
>>>>>>>>
>>>>>>>> I use the  Markowitz model to calculate the optimize some
>>>> assets'
>>>>>>>> weights,
>>>>>>>>
>>>>>>>> and set all the assets' lowerlimit with Zero and set all the
>>>>>>>> upperlimits with
>>>>>>>>
>>>>>>>> One, but it turns out with some negative value which are
>>>> little
>>>>>>>> than Zero, it
>>>>>>>>
>>>>>>>> seems that the lowerlimit does not work.
>>>>>>>>
>>>>>>>> So I want to ask what should I do it to confirm the
>>>> lowerlimit?
>>>>>>>>
>>>>>>>> --
>>>>>>>> Best regards
>>>>>>>>
>>>>>>>> weichao
>>>>>>>>
>>>>>>>>
>>>>>>
>>>> ------------------------------------------------------------------- 
>>>> ---
>>>>>>>> --------
>>>>>>>> This SF.net email is sponsored by:
>>>>>>>> SourcForge Community
>>>>>>>> SourceForge wants to tell your story.
>>>>>>>> http://p.sf.net/sfu/sf-spreadtheword
>>>>>>>> _______________________________________________
>>>>>>>> ojAlgo-user mailing list
>>>>>>>> ojA...@li...
>>>>>>>> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>
>>>> ------------------------------------------------------------------- 
>>>> ---
>>>>>> ------
>>>>>> --
>>>>>>> This SF.net email is sponsored by:
>>>>>>> SourcForge Community
>>>>>>> SourceForge wants to tell your story.
>>>>>>> http://p.sf.net/sfu/sf-spreadtheword
>>>>>>> _______________________________________________
>>>>>>> ojAlgo-user mailing list
>>>>>>> ojA...@li...
>>>>>>> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
>>>>>>
>>>>>>
>>>>>> --
>>>>>> Best regards
>>>>>>
>>>>>>
>>>> ------------------------------------------------------------------- 
>>>> ---
>>>>>> --------
>>>>>> This SF.net email is sponsored by:
>>>>>> SourcForge Community
>>>>>> SourceForge wants to tell your story.
>>>>>> http://p.sf.net/sfu/sf-
>>>>>> spreadtheword_______________________________________________
>>>>>> ojAlgo-user mailing list
>>>>>> ojA...@li...
>>>>>> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
>>>>>
>>>>>
>>>> ------------------------------------------------------------------- 
>>>> ---
>>>> ------
>>>> -- 
>>>>> This SF.net email is sponsored by:
>>>>> SourcForge Community
>>>>> SourceForge wants to tell your story.
>>>>> http://p.sf.net/sfu/sf-spreadtheword
>>>>> _______________________________________________
>>>>> ojAlgo-user mailing list
>>>>> ojA...@li...
>>>>> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
>>>>
>>>>
>>>> -- 
>>>> Best regards
>>>>
>>>> ------------------------------------------------------------------- 
>>>> ---
>>>> --------
>>>> This SF.net email is sponsored by:
>>>> SourcForge Community
>>>> SourceForge wants to tell your story.
>>>> http://p.sf.net/sfu/sf-
>>>> spreadtheword_______________________________________________
>>>> ojAlgo-user mailing list
>>>> ojA...@li...
>>>> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
>>>
>>> -------------------------------------------------------------------- 
>>> --------
>> --
>>> This SF.net email is sponsored by:
>>> SourcForge Community
>>> SourceForge wants to tell your story.
>>> http://p.sf.net/sfu/sf-spreadtheword
>>> _______________________________________________
>>> ojAlgo-user mailing list
>>> ojA...@li...
>>> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
>>
>>
>> --
>> Best regards
>>
>>
>> --------------------------------------------------------------------- 
>> ---------
>> This SF.net email is sponsored by:
>> SourcForge Community
>> SourceForge wants to tell your story.
>> http://p.sf.net/sfu/sf-spreadtheword
>> _______________________________________________
>> ojAlgo-user mailing list
>> ojA...@li...
>> https://lists.sourceforge.net/lists/listinfo/ojalgo-user
>