[R] Linear Regressions with constraint coefficients

Aleksandrovic, Aljosa (Pfaeffikon) Aljosa.Aleksandrovic at man.com
Thu Apr 28 14:42:44 CEST 2016


Hi Gabor,

Thanks a lot for your help!

I tried to implement your nonlinear least squares solver on my data set. I was just wondering about the argument start. If I would like to force all my coefficients to be inside an interval, let’s say, between 0 and 1, what kind of starting values are normally recommended for the start argument (e.g. Using a 4 factor model with b1, b2, b3 and b4, I tried start = list(b1 = 0.5, b2 = 0.5, b3 = 0.5, b4 = 0.5))? I also tried other starting values ... Hence, the outputs are very sensitive to that start argument?     

Thanks a lot for your answer in advance!

Kind regards,
Aljosa



Aljosa Aleksandrovic, FRM, CAIA
Quantitative Analyst - Convertibles
aljosa.aleksandrovic at man.com
Tel +41 55 417 76 03

Man Investments (CH) AG
Huobstrasse 3 | 8808 Pfäffikon SZ | Switzerland

-----Original Message-----
From: Gabor Grothendieck [mailto:ggrothendieck at gmail.com] 
Sent: Dienstag, 26. April 2016 17:59
To: Aleksandrovic, Aljosa (Pfaeffikon)
Cc: r-help at r-project.org
Subject: Re: [R] Linear Regressions with constraint coefficients

This is a quadratic programming problem that you can solve using either a quadratic programming solver with constraints or a general nonlinear solver with constraints.  See https://cran.r-project.org/web/views/Optimization.html
for more info on what is available.

Here is an example using a nonlinear least squares solver and non-negative bound constraints. The constraint that the coefficients sum to 1 is implied by dividing them by their sum and then dividing the coefficients found by their sum at the end:

# test data
set.seed(123)
n <- 1000
X1 <- rnorm(n)
X2 <- rnorm(n)
X3 <- rnorm(n)
Y <- .2 * X1 + .3 * X2 + .5 * X3 + rnorm(n)

# fit
library(nlmrt)
fm <- nlxb(Y ~ (b1 * X1 + b2 * X2 + b3 * X3)/(b1 + b2 + b3),
     data = list(Y = Y, X1 = X1, X2 = X2, X3 = X3),
     lower = numeric(3),
     start = list(b1 = 1, b2 = 2, b3 = 3))

giving the following non-negative coefficients which sum to 1 that are reasonably close to the true values of 0.2, 0.3 and 0.5:

> fm$coefficients / sum(fm$coefficients)
     b1      b2      b3
0.18463 0.27887 0.53650


On Tue, Apr 26, 2016 at 8:39 AM, Aleksandrovic, Aljosa (Pfaeffikon) <Aljosa.Aleksandrovic at man.com> wrote:
> Hi all,
>
> I hope you are doing well?
>
> I’m currently using the lm() function from the package stats to fit linear multifactor regressions.
>
> Unfortunately, I didn’t yet find a way to fit linear multifactor regressions with constraint coefficients? I would like the slope coefficients to be all inside an interval, let’s say, between 0 and 1. Further, if possible, the slope coefficients should add up to 1.
>
> Is there an elegant and not too complicated way to do such a constraint regression estimation in R?
>
> I would very much appreciate if you could help me with my issue?
>
> Thanks a lot in advance and kind regards, Aljosa Aleksandrovic
>
>
>
> Aljosa Aleksandrovic, FRM, CAIA
> Quantitative Analyst - Convertibles
> aljosa.aleksandrovic at man.com
> Tel +41 55 417 7603
>
> Man Investments (CH) AG
> Huobstrasse 3 | 8808 Pfäffikon SZ | Switzerland
>
>
> -----Original Message-----
> From: Kevin E. Thorpe [mailto:kevin.thorpe at utoronto.ca]
> Sent: Dienstag, 26. April 2016 14:35
> To: Aleksandrovic, Aljosa (Pfaeffikon)
> Subject: Re: Linear Regressions with constraint coefficients
>
> You need to send it to r-help at r-project.org however.
>
> Kevin
>
> On 04/26/2016 08:32 AM, Aleksandrovic, Aljosa (Pfaeffikon) wrote:
>> Ok, will do! Thx a lot!
>>
>> Please find below my request:
>>
>> Hi all,
>>
>> I hope you are doing well?
>>
>> I’m currently using the lm() function from the package stats to fit linear multifactor regressions.
>>
>> Unfortunately, I didn’t yet find a way to fit linear multifactor regressions with constraint coefficients? I would like the slope coefficients to be all inside an interval, let’s say, between 0 and 1. Further, if possible, the slope coefficients should add up to 1.
>>
>> Is there an elegant and not too complicated way to do such a constraint regression estimation in R?
>>
>> I would very much appreciate if you could help me with my issue?
>>
>> Thanks a lot in advance and kind regards, Aljosa Aleksandrovic
>>
>>
>>
>> Aljosa Aleksandrovic, FRM, CAIA
>> Quantitative Analyst - Convertibles
>> aljosa.aleksandrovic at man.com
>> Tel +41 55 417 7603
>>
>> Man Investments (CH) AG
>> Huobstrasse 3 | 8808 Pfäffikon SZ | Switzerland
>>
>>
>> -----Original Message-----
>> From: Kevin E. Thorpe [mailto:kevin.thorpe at utoronto.ca]
>> Sent: Dienstag, 26. April 2016 14:28
>> To: Aleksandrovic, Aljosa (Pfaeffikon); r-help-owner at r-project.org
>> Subject: Re: Linear Regressions with constraint coefficients
>>
>> I believe I approved a message with such a subject. Perhaps there was another layer that subsequently rejected it after that. I didn't notice any unusual content. Try again, making sure you send the message in plain text only.
>>
>> Kevin
>>
>> On 04/26/2016 08:16 AM, Aleksandrovic, Aljosa (Pfaeffikon) wrote:
>>> Do you know where I get help for my issue?
>>>
>>> Thanks in advance and kind regards,
>>> Aljosa
>>>
>>>
>>> Aljosa Aleksandrovic, FRM, CAIA
>>> Quantitative Analyst - Convertibles
>>> aljosa.aleksandrovic at man.com
>>> Tel +41 55 417 7603
>>>
>>> Man Investments (CH) AG
>>> Huobstrasse 3 | 8808 Pfäffikon SZ | Switzerland
>>>
>>> -----Original Message-----
>>> From: R-help [mailto:r-help-bounces at r-project.org] On Behalf Of
>>> r-help-owner at r-project.org
>>> Sent: Dienstag, 26. April 2016 14:10
>>> To: Aleksandrovic, Aljosa (Pfaeffikon)
>>> Subject: Linear Regressions with constraint coefficients
>>>
>>> The message's content type was not explicitly allowed
>>>
>
>
> --
> Kevin E. Thorpe
> Head of Biostatistics,  Applied Health Research Centre (AHRC)
> Li Ka Shing Knowledge Institute of St. Michael's Hospital
> Assistant Professor, Dalla Lana School of Public Health
> University of Toronto
> email: kevin.thorpe at utoronto.ca  Tel: 416.864.5776  Fax: 416.864.3016
>
> This email has been sent by a member of the Man group (“Man”). Man’s parent company, Man Group plc, is registered in England and Wales (company number 08172396) at Riverbank House, 2 Swan  Lane, London, EC4R 3AD.
> The contents of this email are for the named addressee(s) only. It contains information which may be confidential and privileged. If you are not the intended recipient, please notify the sender immediately, destroy this email and any attachments and do not otherwise disclose or use them. Email transmission is not a secure method of communication and Man cannot accept responsibility for the completeness or accuracy of this email or any attachments. Whilst Man makes every effort to keep its network free from viruses, it does not accept responsibility for any computer virus which might be transferred by way of this email or any attachments. This email does not constitute a request, offer, recommendation or solicitation of any kind to buy, subscribe, sell or redeem any investment instruments or to perform other such transactions of any kind. Man reserves the right to monitor, record and retain all electronic and telephone communications through its network in accordance with applicable laws and regulations. --UwQe9f5k7pI3vplngP
>
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.



-- 
Statistics & Software Consulting
GKX Group, GKX Associates Inc.
tel: 1-877-GKX-GROUP
email: ggrothendieck at gmail.com


More information about the R-help mailing list