R help - Feb 2009 - Generating Numbers With Certain Distribution in R

Dear all,

Is there a way to generate K numbers of integer (K = 10^6).
The maximum value of the integer is 200,000 and minimum is 1.

And the occurrences of this integer follows
a lognormal distribution.

- Gundala Viswanath
Jakarta - Indonesia

Gundala Viswanath <gundalav <at> gmail.com> writes:
> 
> Dear all,
> 
> Is there a way to generate K numbers of integer (K = 10^6).
> The maximum value of the integer is 200,000 and minimum is 1.
> 
> And the occurrences of this integer follows
> a lognormal distribution.
> 
> - Gundala Viswanath
> Jakarta - Indonesia
  Technically speaking, I don't think this is possible
since lognormal variates range from 0 to infinity
and are continuous.  However, perhaps something like

x <- rlnorm(1e6,mulog=1,sdlog=1) ## pick any parameters you like
y <- round((x-min(x))/diff(range(x)))*199999+1

  That probably doesn't do exactly what you want,
but you will probably need to specify your problem more
precisely (and perhaps give some context) if you want
a better answer ...

  Ben Bolker

On Wed, Feb 11, 2009 at 2:15 PM, Ben Bolker <bolker at ufl.edu>
wrote:
> Bernardo Rangel Tura wrote:
>
>> I think your routine need a little fix
>>
>> x <- rlnorm(1e6,meanlog=1,sdlog=1) ## pick any parameters you like
>> y <- round((x-min(x)/diff(range(x)))*199999+1)
>>
>> What you think?
>
>  Yes.

No.
Bernardo misplaced the parenthesis around (x-min(x))
Correct version is:

x <- rlnorm(1e6,meanlog=1,sdlog=1) ## pick any parameters you like
y <- round((x-min(x))/diff(range(x))*199999+1)


/Gustaf


-- 
Gustaf Rydevik, M.Sci.
tel: +46(0)703 051 451
address:Essingetorget 40,112 66 Stockholm, SE
skype:gustaf_rydevik

Dear all,
Thanks so much, the latest version by Gustav work just fine.

In the first version, the outcome is more digital
than contiguous.

- Gundala Viswanath
Jakarta - Indonesia



On Wed, Feb 11, 2009 at 10:35 PM, Gustaf Rydevik
<gustaf.rydevik at gmail.com> wrote:
> On Wed, Feb 11, 2009 at 2:15 PM, Ben Bolker <bolker at ufl.edu>
wrote:
>> Bernardo Rangel Tura wrote:
>>
>>> I think your routine need a little fix
>>>
>>> x <- rlnorm(1e6,meanlog=1,sdlog=1) ## pick any parameters you
like
>>> y <- round((x-min(x)/diff(range(x)))*199999+1)
>>>
>>> What you think?
>>
>>  Yes.
>
>
> No.
> Bernardo misplaced the parenthesis around (x-min(x))
> Correct version is:
>
> x <- rlnorm(1e6,meanlog=1,sdlog=1) ## pick any parameters you like
> y <- round((x-min(x))/diff(range(x))*199999+1)
>
>
> /Gustaf
>
>
> --
> Gustaf Rydevik, M.Sci.
> tel: +46(0)703 051 451
> address:Essingetorget 40,112 66 Stockholm, SE
> skype:gustaf_rydevik
>