this is indeed a strange request for 2 reasons:
1. a mean of 3.5 surely is not an integer, right?
2. with a limited range and only about 6 DISCRETE values ... then it
certainly is NOT a normal distribution
but, i did a simulation in minitab below ...
MTB > rand 1000 c1;
SUBC> norm 3 1.
MTB > roun c1 c2 <<<<< note ... i rounded the values to get integers
MTB > dotp c2
Dotplot: C2
Each dot represents up to 29 points
.
:
:
. :
: : :
: : :
. : : :
. . : : : : : .
-+---------+---------+---------+---------+---------+-----C2
-1.5 0.0 1.5 3.0 4.5 6.0
MTB > desc c2
Descriptive Statistics: C2
Variable N Mean Median TrMean StDev SE Mean
C2 1000 2.9250 3.0000 2.9200 1.0404 0.0329
Variable Minimum Maximum Q1 Q3
C2 -1.0000 6.0000 2.0000 4.0000
MTB >
now here is the NON integer version of the data ...
Each dot represents up to 4 points
. . .
.:.:::.
..::::::: .
:::::::::::
..:::::::::::
:::::::::::::: :
..::::::::::::::.:::.
. .::::::::::::::::::::::::
.. :.:::::::::::::::::::::::::::::....... .
+---------+---------+---------+---------+---------+-------C1
-1.5 0.0 1.5 3.0 4.5 6.0
dotp c1
At 01:11 AM 8/17/00 -0400, Chris Chiu wrote:
>Hello:
>
>Does anyone happen to know of a way to generate integers that have a
>normal distribution or a distribution close to a normal distribution?
>
>I am trying to generate some whole-numbers with a given range, mean, and
>standard deviation. For example, I was wondering if one could simulate,
>say, 1000 random numbers with a mean 3.5, standard deviation 1, and a
>range going from 1 to 6?
>
>Any comments will be greatly appreciated. Thanks
>
>Sincerely,
>Chris
>
>
>
>=================================================================
>Instructions for joining and leaving this list and remarks about
>the problem of INAPPROPRIATE MESSAGES are available at
> http://jse.stat.ncsu.edu/
>=================================================================
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
