Subscribe Bookmark RSS Feed

JMP Script: Random Functions

clausa

Community Trekker

Joined:

Jan 16, 2014

Hi Folks,

I am trying to use a variety of random functions for continuous plots. I find that many exist in JSL, but many do not. Am I missing something here? If you look below you will see in blue the expressions that exist, while those in black do not and I list the variables that I would expect them to take.

Thanks

random normal(mu, sigma);

random log normal(mu, sigma);

random weibull(beta, alpha);     //flipflopped ==> alpha, beta

random weibull threshold;           //alpha, beta, theta

random extreme vaule();               //lamda, delta

random exponential ();                 //sigma

random gamma (lamda, scale);     //not enough variables

ramdom beta ();                               //alpha, beta, theta, sigma

random normal mixture([meanvec],[sdvec],[probvec]);

random johnson su(gamma, delta, theta, sigma);

random johnson sb(gamma, delta, theta, sigma);

random johnson sl(gamma, delta, theta, sigma);

random glog(mu, sigma, lambda);

1 ACCEPTED SOLUTION

Accepted Solutions
ms

Super User

Joined:

Jun 23, 2011

Solution

The available random functions are listed in the Scripting Index (Help Menu).

I found these in JMP 10. There is a bunch of new functions in version 11.

Eval List({

  Random Beta( 1, 1, 0, 1 ),

  Random LEV( 10, 5 ), // Extreme value dist.

  Random Exp() * 2, // Exponential, λ = 2

  Random Weibull( 3, 20 ) + 5  //Theta = 5

  }

)


{0.363363575496819, 6.33214952243719, 1.30944855321776, 13.3691613532646}

2 REPLIES
ms

Super User

Joined:

Jun 23, 2011

Solution

The available random functions are listed in the Scripting Index (Help Menu).

I found these in JMP 10. There is a bunch of new functions in version 11.

Eval List({

  Random Beta( 1, 1, 0, 1 ),

  Random LEV( 10, 5 ), // Extreme value dist.

  Random Exp() * 2, // Exponential, λ = 2

  Random Weibull( 3, 20 ) + 5  //Theta = 5

  }

)


{0.363363575496819, 6.33214952243719, 1.30944855321776, 13.3691613532646}

clausa

Community Trekker

Joined:

Jan 16, 2014

Thanks MS. I totally just mistyped the Beta one. For the others that makes sense.

Do you know with the LEV how that correlates to the Lambda and Delta values that come from JMP distribution analysis? JSL asks for random LEV(mu, signma)

4908_Extreme Value Fit Pic.jpg

I tried running them as just the mu and sigma that JSL requests, but that does not work. I cannot find any connections online, but I do not know these statistics themselves that well.

For Gamma, a 500,000 point simulation suggests that the lambda that JSL asks for random gamma (lamda, scale)is the alpha from the estimate and the scale is the scale/sigma, but can anyone definitively confirm? I assume threshold is just a shift at the end (e.g +0 for the example below).

4909_Gamma Fit Pic.jpg

Why does JMP sometimes use the same letters and/or names, but others times not. Lambda vs alpha, scale and scale, scale vs sigma.

Thanks

Message was edited by: Andrew Claus