Thread Subject:

Selecting a "random" number based on second row of matrix and other criteria

I have something like the following:

A = [1 2 3 4 5 6; 9 5 7 1 2 6; 6 2 3 1 4 5];

Max = 8;

I am looking to be able to choose a "random" number that is contained in the second column. Therefore the number can be 9 5 7 1 2 or 6 and also be less than the Max = 8;

So, a number to be chosen that is in the list of 1 2 5 6 7 (for this case).

Is this feasible?

On 10/13/2012 4:15 PM, Kenny wrote:
> I have something like the following:
>
> A = [1 2 3 4 5 6;
        9 5 7 1 2 6;
        6 2 3 1 4 5];
>
> Max = 8;
>
> I am looking to be able to choose a "random" number that is contained in the second column.
>Therefore the number can be 9 5 7 1 2 or 6 and also be less than the Max = 8;
>

How can the number be '9' and also be less than Max=8 at the same time? btw, the
second column has {2,5,6} numbers in it. What you typed above is the second
row?

> So, a number to be chosen that is in the list of 1 2 5 6 7 (for this case).
>
> Is this feasible?
>

to select a random number from a vector, one way

------------------------------
EDU>> v=[1 2 3 4 5 6];
EDU>> v(randi(length(v)))
------------------------------

or

------------------------------
EDU>> v=[1 2 3 4 5 6];
EDU>> r=randperm(length(v));
EDU>> v(r(1))
-------------------------------

so for your case, your the second column (or the second row) in place
of v above.

--Nasser

"Nasser M. Abbasi" <nma@12000.org> wrote in message <k5cm3v$ga$1@speranza.aioe.org>...
> On 10/13/2012 4:15 PM, Kenny wrote:
> > I have something like the following:
> >
> > A = [1 2 3 4 5 6;
> 9 5 7 1 2 6;
> 6 2 3 1 4 5];
> >
> > Max = 8;
> >
> > I am looking to be able to choose a "random" number that is contained in the second column.
> >Therefore the number can be 9 5 7 1 2 or 6 and also be less than the Max = 8;
> >
>
> How can the number be '9' and also be less than Max=8 at the same time? btw, the
> second column has {2,5,6} numbers in it. What you typed above is the second
> row?
>
> > So, a number to be chosen that is in the list of 1 2 5 6 7 (for this case).
> >
> > Is this feasible?
> >
>
> to select a random number from a vector, one way
>
> ------------------------------
> EDU>> v=[1 2 3 4 5 6];
> EDU>> v(randi(length(v)))
> ------------------------------
>
> or
>
> ------------------------------
> EDU>> v=[1 2 3 4 5 6];
> EDU>> r=randperm(length(v));
> EDU>> v(r(1))
> -------------------------------
>
> so for your case, your the second column (or the second row) in place
> of v above.
>
> --Nasser
>
>
>
>
>
Great, thanks! Would it be possible to integrate the Max, so the random number is less than or equal to the Max, in addition to being one of those numbers?

For your example:

clear all;

 v=[1 2 3 4 5 6];
 v
 Max = 5;
 
 g=v(randi(length(v)));
 g

basically adding to it that the random number cannot be greater than the Max.

On 10/13/2012 4:42 PM, Kenny wrote:

> Great, thanks! Would it be possible to integrate the Max, so the
> random number is less than or equal to the Max, in addition to being one of those numbers?
>
> For your example:
>
> clear all;
>
> v=[1 2 3 4 5 6];
> v
> Max = 5;
>
> g=v(randi(length(v)));
> g
>
> basically adding to it that the random number cannot be greater than the Max.
>

sure. Simply filter out the data you do not want first, then apply
the same logic.

To filter out everything below max, use find() then apply randi() to
the resulting vector.

-----------------
v = [1 2 3 4 5 6 7 8];
myMax = 5;
tmp = v(v<myMax)
tmp(randi(length(tmp)))
-----------------------

"Nasser M. Abbasi" <nma@12000.org> wrote in message <k5cpkd$8h6$1@speranza.aioe.org>...
> On 10/13/2012 4:42 PM, Kenny wrote:
>
> > Great, thanks! Would it be possible to integrate the Max, so the
> > random number is less than or equal to the Max, in addition to being one of those numbers?
> >
> > For your example:
> >
> > clear all;
> >
> > v=[1 2 3 4 5 6];
> > v
> > Max = 5;
> >
> > g=v(randi(length(v)));
> > g
> >
> > basically adding to it that the random number cannot be greater than the Max.
> >
>
> sure. Simply filter out the data you do not want first, then apply
> the same logic.
>
> To filter out everything below max, use find() then apply randi() to
> the resulting vector.
>
> -----------------
> v = [1 2 3 4 5 6 7 8];
> myMax = 5;
> tmp = v(v<myMax)
> tmp(randi(length(tmp)))
> -----------------------
>
Thank you. I was thinking in the reverse -- filtering out the data first makes sense.