random selection

[kutya 0]

I want to use random selection from an array.

Let's have:

obj = [obj1, obj2, obj3, obj4]

I saw in a script the next:

<table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Code Sample </td></tr><tr><td id="CODE">objs select (random ((count objs) - 1))}

As I know random returns real.

It's not clear to me how can you select with a real number?

Only have two ideas: the random number is truncated, or rounded. Does someone know which is? I think trunc, but not sure.

[ProfTournesol 956]

Well, an easy way to round a random number is using "mod" command :

<table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Code Sample </td></tr><tr><td id="CODE">

_r = random 2

; this return a real value between 0 and 2

_t = _r - (_r mod 1)

; this return a rounded value of the previous real number "_r"

[kutya 0]

Thanks for the info!

[Metal Heart 0]

Well, an easy way to round a random number is using "mod" command :

<table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Code Sample </td></tr><tr><td id="CODE">

_r = random 2

; this return a real value between 0 and 2

_t = _r - (_r mod 1)

; this return a rounded value of the previous real number "_r"

Odds for getting 2 out of that is one (2.0) against two million (0.0-1.999999).

[kutya 0]

Yeah, I noticed that, that's why I thought it's better to use trunc. But didn't test it yet. In that case, the odds would be the same.

[Chris Death 0]

errm - forget what i wrote - i can't test it right now because

i'm at work and can't test anything to proove it.

And i mixed up something during reading the post.

~S~ CD

[kutya 0]

After implementing, the command in the first post really works! It's made by Silola. He's really a genius. This way, if you define an array, you can in a single line choose a random element. If you need some random trucks in a convoy, define an array with truck model names, and a couple of times call createvehicle with the above mentioned code, and you've got a convoy with random trucks. Really simplifies the lengthy case structure.

[Metal Heart 0]

Actually it's flawed. The first and last array element have only half the chance of happening than the ones in the middle.

<table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Code Sample </td></tr><tr><td id="CODE">arr = [0,1,2,3];

arr select (random ((count arr) - 1)); // random 3 in this case

Here are the ranges:

element 0 = 0.0 - 0.499999

element 1 = 0.5 - 1.499999

element 2 = 1.5 - 2.499999

element 3 = 2.5 - 3.0

[kutya 0]

I thought about this, but was too busy to test it. I didn't know, is the select command truncating or rounding if it has a real as a parameter. I would like if it was trunc, but you proved otherwise. With trunc it would work correctly I think. But surely, thanks for the info!

[cleanrock 0]

i do it this way :

_list = [0, 1, 2, 3]

_entry = _list select ((random (count _list)) - 0.5)

this will select a random entry in the list and all entries will have equal probability of being selected

[Chris Death 0]

Well, but then the first element doesn't have half of the

chance.

If you subtract 1 from 0 it's -1

Maybe the chances are even if you subtract only 0.5 from

the whole array count. ((random 4) - 0.5).

Only what i dunno now is if it would produce a minus number

in this case anyway, but if not you could do it with one line

afterwards creating the random.

~S~ CD

[Metal Heart 0]

Quote[/b] ]_list = [0, 1, 2, 3]

_entry = _list select ((random (count _list)) - 0.5)

this will select a random entry in the list and all entries will have equal probability of being selected

It returns null if random hits the maximum value. A very small chance though but it could happen

Quote[/b] ]Maybe the chances are even if you subtract only 0.5 from

the whole array count. ((random 4) - 0.5).

That's exactly what it does, see the parenthesis.

[cleanrock 0]

It returns null if random hits the maximum value. A very small chance though but it could happen

i tried it and this is the way bis do it, if u make a script which test this i am pretty sure it will never be the max value

[Metal Heart 0]

I guess I have to agree if BIS uses such script too. However, here's what the command reference says:

Quote[/b] ]random x

Operand types:

x: Number

Type of returned value:

Number

Description:

random real value from 0 to x

Example:

random 1