Dear my RG friends,
I am going to model the Wind speed in General Algebraic Modeling System (GAMS) software. Since Weibull Distribution Function (WDF) is the best mathematical tool to represent wind speeds, I need to use a GAMS statement in order to generate wind speeds based on WDF.
Is there Weibull Distribution function in GAMS?
I look forward to reading your worthwhile answer.
Regards.
Dear Ms. Sefeedpari,
I highly appreciate your worthwhile answer. I found out that the interface between GAMS and MATLAB is absolutely necessary and helpful for my optimization problem especially when I am going to use special distribution functions such as Weibull.
Regards,
Morteza Shabanzadeh
Dear Morteza,
you can use external functions built in on whatever language you like, for example matlab, and call them from gams. see the maccarl guide it's very descriptive
regards,
juan
Dear Juan,
Many thanks for your prompt answer. I downloaded McCarl User's Guide 2014 but unfortunately it does not have any descriptive part regarding MATLAB and its interface with GAMS. It has just referred to Ferris paper which is not really fruitful to gams users (in my opinion). Because this paper does not have more examples and thus cannot help us to tackle the challenges regarding GAMS-MATLAB interface.
I would be grateful if you give me a simple example code to understand how this interfacing can be done.
In advance, I apologize for any inconvenience.
Regards,
Morteza
http://mortezash.blogfa.com/
this is an example:
$call xlstalk.exe -s Data.xls
$call function Data par_input sheet1 A2:N89 sheet2 A2:A70
execute 'xlstalk.exe -o Data.xls';
that was the *.gms file, but you need to have been built in the executable "function.exe", which i did in matlab.
hope this helps!
Hi Juan,
I am really grateful to you for your worthwhile example. I did not know there is a 'xlstalk.exe' function in GAMS.
Certainly, it will be fruitful for my GAMS-MATLAB code.
All the best,
Morteza
Sorry, may be I didn't understand your question clear... did you mean Weibull CDF ? If "yes", it is easy to make directly:
CDF = 1 - exp(-R*t^K), so random t according Weibull CDF : t = ( - log (1 - rand)/R)^(1/K), where rand in uniform from [0...1]
Dear Sergey,
Many thanks for your worthwhile answer. That is a good method to generate uniform random variable and get Weibull CDF. It can be implemented in GAMS easily.
Best wishes,
Morteza
http://mortezash.blogfa.com/
I fairly suggest you MATLAB in which you can estimate Weibull parameters (c & k) and also calculate CDP and PDF values easily. The Statistics Toolbox help would help you ceratinly. Actually, I 've used it for wind data, as well.
Dear Ms. Sefeedpari,
I highly appreciate your worthwhile answer. I found out that the interface between GAMS and MATLAB is absolutely necessary and helpful for my optimization problem especially when I am going to use special distribution functions such as Weibull.
Regards,
Morteza Shabanzadeh
Dear Morteza,
Now, 2 years after your question, a simple way for do this is as follows:
$funclibin stolib stodclib
set s /s1*s100/;
parameters wd(s);
function randweibull /stolib.dweibull /
setseedh /stolib.SetSeed /;
scalar setseedh_; setseedh_=setseedh(0);
wd(s) = randweibull(1.84, 7.8);
For more info, see the links below.
https://www.gams.com/help/index.jsp?topic=%2Fgams.doc%2Fuserguides%2Fmccarl%2Fprobability_distribution_funct.htm
https://www.gams.com/help/topic/gams.doc/userguides/mccarl/randnumbers.gms
- Badges
- Science topic
- Similar topics
- Mathematics