<>
" With regards to the farthest maximum point that I was getting ,from the
thousands and thousands of draws that I took,that was 6.23026 ,would
that be a lot larger ,to infinitely large,as well ?"
You can explore this question with the simulation below. It will return
5.7297182, so you can view yours as large. You can also increase the number
of repetitions. Still, theoretically any value can be drawn from this
distribution (you may have to wait a couple of million years for it to come
up), so the discussion is indeed a little "fruitless".
*************
//drop it beforehand
capt prog drop sim
set seed 1324543
//define the program
program define sim
version 10.1
drop _all
set obs 10000000
tempvar z
gen `z' = rnormal()
summ `z', mean
end
//simulate it!
simulate max=r(max), /*
*/ reps(10): sim
su, mean
di in red /*
*/ "Maximum of Maximums: " /*
*/ r(max)
*************
HTH
Martin
-----Ursprüngliche Nachricht-----
Von: [email protected]
[mailto:[email protected]] Im Auftrag von [email protected]
Gesendet: Montag, 5. Oktober 2009 12:59
An: [email protected]
Betreff: Re: st: Number of values in Gaussian Normal Distribution
Thank you for your input.If I understand correctly,if the gap,between two
values in sequence,generated by Stata,were a lot smaller,to infinitely
smaller,I would have gotten ,a lot more than 68518 values ,to an
infinite (uncountable) number of values ,for the points between 0 and
00004.
With regards to the farthest maximum point that I was getting ,from the
thousands and thousands of draws that I took,that was 6.23026 ,would
that be a lot larger ,to infinitely large,as well ?
On 4/10/2009, "David Greenberg" <[email protected]> wrote:
>This is a fruitless enterprise. The normal distribution is continuous. It
takes on an infinite number of values even if you restrict yourself to a
part of the distribution that lies between two points. The number of values
is uncountable.
>- David Greenberg, Sociology Department, New York University
>
>----- Original Message -----
>From: "[email protected]" <[email protected]>
>Date: Sunday, October 4, 2009 7:20 pm
>Subject: st: Number of values in Gaussian Normal Distribution
>To: "[email protected]" <[email protected]>
>
>
>> Dear Statlist,
>> I am trying to locate the number of values that constitute a Gaussian
>> Normal Distribution.I am working only on the right side,the positive
>> side,assuming that the left side is the negative of the positive side.
>> The way I am going about concluding the number of values in the right
>> hand side of the Gaussian Normal Distribution,is by taking thousand upon
>> thousands of draws from the Gaussian distribution, and keeping the
>> values for the particular interval,each time accumulating,sorting and
>> dropping the repeating values until I notice that the particular
>> interval doesn't grow any further.
>> I start with the interval (0 to .00004) ,as the first interval and then
>> (.00004 to .00008),as the second interval ..... , ( .0076 to .0078)
>> in
>> the 195th interval ,which seems to have roughly,the same number of
>> values,very close to 68518.I discovered some patterns in the number of
>> values that intervals hold,which made it somewhat easier for me.The
>> following is what I have found so far.Note that the variable x ,is the
>> number of time the same number of values,repeats.t
>>
>> val min max intrv num x tot
>> 68518 0 .00004 .00004 1 0 0
>> 68525 .00776 .0078 .00004 195 195 13362375
>> 50937 .0078 .00784 .00004 196 1 50937
>> 42939 .00784 .00788 .00004 197 0 0
>> 42944 .01556 .0156 .00004 390 194 8331136
>> 34890 .0156 .01564 .00004 391 1 34890
>> 53687 .01564 .01574 .0001 392 0 0
>> 53687 .03114 .03124 .0001 547 156 8375172
>> 13420 .03124 .03128 .00004 548 1 13420
>> 53687 .03128 .03148 .0002 549 0 0
>> 53687 .06228 .06248 .0002 704 156 8375172
>> 18790 .06248 .0626 .00012 705 1 18790
>> 53687 .0626 .063 .0004 706 0 0
>> 53687 .1246 .125 .0004 861 156 8375172
>> 67109 .125 .126 .001 862 0 0
>> 67108 .249 .25 .001 986 125 8388500
>> 33555 .25 .251 .001 987 0 0
>> 33554 .499 .5 .001 1236 250 8388500
>> 33555 .5 .502 .002 1237 0 0
>> 33554 .998 1 .002 1486 250 8388500
>> 33555 1 1.004 .004 1487 0 0
>> 33554 1.996 2 .004 1736 250 8388500
>> 41944 2 2.01 .01 1737 1 41944
>> 41943 2.01 2.02 .01 1738 0 0
>> 41925 3.44 3.45 .01 1881 144 6037200
>> 41927 3.45 3.46 .01 1882 1 41927
>> 41880 3.46 3.47 .01 1883 1 41880
>> 40876 3.47 3.48 .01 1884 1 40876
>> 39480 3.48 3.49 .01 1885 1 39480
>> 38124 3.49 3.5 .01 1886 1 38124
>> 36813 3.5 3.51 .01 1887 1 36813
>> 35541 3.51 3.52 .01 1888 1 35541
>> 34307 3.52 3.53 .01 1889 1 34307
>> 33126 3.53 3.54 .01 1890 1 33126
>> 31980 3.54 3.55 .01 1891 1 31980
>> 30846 3.55 3.56 .01 1892 1 30846
>> 29779 3.56 3.57 .01 1893 1 29779
>> 28723 3.57 3.58 .01 1894 1 28723
>> 27718 3.58 3.59 .01 1895 1 27718
>> 26748 3.59 3.6 .01 1896 1 26748
>> 25797 3.6 3.61 .01 1897 1 25797
>> 24885 3.61 3.62 .01 1898 1 24885
>> 24001 3.62 3.63 .01 1899 1 24001
>> 23140 3.63 3.64 .01 1900 1 23140
>> 22316 3.64 3.65 .01 1901 1 22316
>> 21516 3.65 3.66 .01 1902 1 21516
>> 20743 3.66 3.67 .01 1903 1 20743
>> 19998 3.67 3.68 .01 1904 1 19998
>> 19270 3.68 3.69 .01 1905 1 19270
>> 18569 3.69 3.7 .01 1906 1 18569
>> 17902 3.7 3.71 .01 1907 1 17902
>> 17250 3.71 3.72 .01 1908 1 17250
>> 16623 3.72 3.73 .01 1909 1 16623
>> 15994 3.73 3.74 .01 1910 1 15994
>> 15394 3.74 3.75 .01 1911 1 15394
>> 14826 3.75 3.76 .01 1912 1 14826
>> 14255 3.76 3.77 .01 1913 1 14255
>> 13738 3.77 3.78 .01 1914 1 13738
>> 13235 3.78 3.79 .01 1915 1 13235
>> 12605 3.79 3.78 .01 1916 1 12605
>> 38335 3.78 3.81 .03 1917 1 38335
>> 44761 3.81 3.85 .04 1918 1 44761
>> 47063 3.85 3.9 .05 1919 1 47063
>> 38734 3.9 3.95 .05 1920 1 38734
>> 31780 3.95 4 .05 1921 1 31780
>> 47278 4 4.1 .1 1922 1 47278
>> 52029 4.1 4.3 .2 1923 1 52029
>> 36661 4.3 6.3 2 1924 1 36661
>>
>> According to the previous data,the number of values that make up the
>> Gaussian distribution is 87796774 * 2 = 175593548.I am wondering if
>> there is a simpler way of calculating the number of values,that
>> constitutes the Gaussian Distribution.
>> Vicror M. Zammit
>>
>> *
>> * For searches and help try:
>> * http://www.stata.com/help.cgi?search
>> * http://www.stata.com/support/statalist/faq
>> * http://www.ats.ucla.edu/stat/stata/
>*
>* For searches and help try:
>* http://www.stata.com/help.cgi?search
>* http://www.stata.com/support/statalist/faq
>* http://www.ats.ucla.edu/stat/stata/
>
>
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
