Best Optimization Routines

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

Best Optimization Routines

JamesRyan(Ryan James)/E&P North America Business Division
I have a set of data, the production of oil from a well.  And an equation to predict that forecast. 

The equation requires 5 input variables which are real numbers with upper and lower bounds, 1 input variable which must be an integer and 1 input variable which can be 1 of 2 string variables. 

What is the best optimization routine (I am using nlminb) to use in R to determine the best set of inputs and how to I handle the integer and string inputs when optimizing?  

I have attached a plot of what I am trying to accomplish as well as my sample code.  

Thanks,

Ryan



______________________________________________
[hidden email] 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.

Rplot.pdf (5K) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Best Optimization Routines

paul smith-6
On Tue, Feb 13, 2018 at 7:55 PM, JamesRyan(Ryan James)/E&P North
America Business Division <[hidden email]> wrote:
> I have a set of data, the production of oil from a well.  And an equation to predict that forecast.
>
> The equation requires 5 input variables which are real numbers with upper and lower bounds, 1 input variable which must be an integer and 1 input variable which can be 1 of 2 string variables.
>
> What is the best optimization routine (I am using nlminb) to use in R to determine the best set of inputs and how to I handle the integer and string inputs when optimizing?
>
> I have attached a plot of what I am trying to accomplish as well as my sample code.

If I understand correctly your problem, your equation is linear, and
you have continuous and integer variables (the string input can be
coded as a binary integer variable). Thus, you can use a package to
deal with mixed integer programming:

http://blog.revolutionanalytics.com/2016/12/mixed-integer-programming-in-r-with-the-ompr-package.html

Hope this helps you,

Paul

______________________________________________
[hidden email] 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.
Reply | Threaded
Open this post in threaded view
|

Re: Best Optimization Routines

JamesRyan(Ryan James)/E&P North America Business Division
Thanks Paul.  Appreciate the help.  I'll give this a shot.  

Ryan James
D: 713.395.1794
M: 214.843.7301

-----Original Message-----
From: Paul Smith [mailto:[hidden email]]
Sent: Thursday, February 15, 2018 7:12 AM
To: Ryan James/E&P North America Business Division/SKI
Cc: [hidden email]
Subject: Re: [R] Best Optimization Routines

On Tue, Feb 13, 2018 at 7:55 PM, JamesRyan(Ryan James)/E&P North America Business Division <[hidden email]> wrote:
> I have a set of data, the production of oil from a well.  And an equation to predict that forecast.
>
> The equation requires 5 input variables which are real numbers with upper and lower bounds, 1 input variable which must be an integer and 1 input variable which can be 1 of 2 string variables.
>
> What is the best optimization routine (I am using nlminb) to use in R to determine the best set of inputs and how to I handle the integer and string inputs when optimizing?
>
> I have attached a plot of what I am trying to accomplish as well as my sample code.

If I understand correctly your problem, your equation is linear, and you have continuous and integer variables (the string input can be coded as a binary integer variable). Thus, you can use a package to deal with mixed integer programming:

http://blog.revolutionanalytics.com/2016/12/mixed-integer-programming-in-r-with-the-ompr-package.html

Hope this helps you,

Paul
______________________________________________
[hidden email] 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.