Random Number Generation in a given range

[Guillermo Estrada]

Hi, I've been reading some posts but I still have the same question.

I need to generate a random number between a min and a max value of floats (Float64) for example:

random(-0.001, 0.001)

Is there any way pkg "math/rand" could do this using the Normal Distribution or such? Most Random generation on the pkg works in [0.0,1.0)

The closest I found is:

func (*Rand) NormFloat64

But I need a way to clamp that range into my own, any ideas? Thnx in advance!

[Andrew Gerrand]

A normal distribution has no upper or lower bound, so NormFloat64
might not be what you want.

This expression will give you a random float bounded to [-0.001,0.001):

(rand.Float64() * 2 - 1) * 0.001

Andrew

[si guy]

If you clamp a normal distribution then it isn't a normal distribution anymore... Are you saying that you want [-0.001,0.001] to be within n sigma of a mean of zero?

If that's the case then
MyVal := rand.NormFloat64() * ( 0.001 / n )

You can clamp it afterwards with a switch
switch {
case MyVal > 0.001:
MyVal = 0.001
case MyVal < -0.001:
MyVal = -0.001
}

But again, this won't be a normal distribution anymore...

[Guillermo Estrada]

A normal distribution has no upper or lower bound, so NormFloat64
might not be what you want.

This expression will give you a random float bounded to [-0.001,0.001):

    (rand.Float64() * 2 - 1) * 0.001

Andrew

This expression seems pretty nice and compact. It solves my specific problem right now. I have been analyzing it and it works with any range where max min values are on negative and positive sides respectively and are of course the same.

How can it be done to use any arbitrary max min values?

ie. 

random(23.45, 54.8)

random(-1.123, 987.23)

random(-82.12,-2.65)

Of course I know this will need a lot more coding, but I use a lot of random numbers in computer graphic algorithms and it will be best to have a working function.

Thanx a lot Andrew, my brain was kinda dry already :P

[Matt Kane's Brain]

This should do it:

func random(min, max float64) float64 {
return rand.Float64() * (max - min) + min
}

On Tue, Mar 27, 2012 at 20:49, Guillermo Estrada <phro...@gmail.com> wrote:
> How can it be done to use any arbitrary max min values?

--
matt kane's brain
http://hydrogenproject.com

[Guillermo Estrada]

This should do it:

func random(min, max float64) float64 {
  return rand.Float64() * (max - min) + min
}

It truly does as far as I can tell :D thanks a lot! My brain can rest in peace now! In my road to learning Go, I ported a raytracer (I'm impressed with Go's speed) and now I'm making a web application that generate deJong Attractors using density maps. The more I use Go the more caveats I found, but at the same time the more I like it. And the cream of the crop of this language is truly the community behind it!

[alok....@practo.com]

A python like range utility which can be used for this

https://github.com/alok87/goutils/blob/master/pkg/random/random.go

```random.RangeInt(2, 100, 3)```

[peterGo]

Alok,

Your import path "github.com/alok87/goutils/random" is not valid for "go get". The source is at "github.com/alok87/goutils/pkg/random".

You write "for { rand.Seed(); rand.Intn(); }". Therefore, the range of values is not random.

You write "arr[r] = rand.Intn(max) + min" which is incorrect.

And so on.

Peter

[Alok Kumar Singh]

I have fixed it thanks

--
You received this message because you are subscribed to a topic in the Google Groups "golang-nuts" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/golang-nuts/_M-8hRpQs84/unsubscribe.
To unsubscribe from this group and all its topics, send an email to golang-nuts+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

Regards,

Alok Kumar Singh 

Platform Engineering

+ 91 9880121029 

[Sebastien Binet]

Alok,

On Mon, Jul 10, 2017 at 3:04 PM, Alok Kumar Singh <alok....@practo.com> wrote:

I have fixed it thanks

I wouldn't recommend using rand.Seed as you do: this defeats any attempt at getting reproductible results.

-s

[Alok Kumar Singh]

in my usecasei use this to generate random slice everytime 

dont want the same result

using it to test my algos

[Jan Mercl]

On Mon, Jul 10, 2017 at 4:41 PM Alok Kumar Singh <alok....@practo.com> wrote:

> in my usecasei use this to generate random slice everytime
> dont want the same result
> using it to test my algos

For that you need one only call to Seed at init().

--

-j

[alok....@practo.com]

yep you are right ! fixed it thanks 

https://github.com/alok87/goutils/commit/9f5f19a65eedf1bc9be82ec3f3b42c92142b2b8e