Jump to content
DOSBODS
  • Welcome to DOSBODS

     

    DOSBODS is free of any advertising.

    Ads are annoying, and - increasingly - advertising companies limit free speech online. DOSBODS Forums are completely free to use. Please create a free account to be able to access all the features of the DOSBODS community. It only takes 20 seconds!

     

Poisson distribution and SD%


Recommended Posts

Hi Everyone,

not sure if we have any stats experts out there but I have been trying to Google this without any success...

Now with a Normal distribution 1 standard deviation (SD) about for 68.3% from the mean, 2SD=95.4% and 3SD=99.7%, is this the same for a Poisson distribution and/or other parametric distributions. Also, as the mean increases in the Poisson distribution its shape changes so is the above SD% the same for all  Poisson distribution shapes OR just when it becomes sufficient to be 'Normal' like i.e mean = > ~15?

Thanks for any help, if you could include a link/source as well that would be great!

Link to post
Share on other sites
4 hours ago, MrXxxx said:

Hi Everyone,

not sure if we have any stats experts out there but I have been trying to Google this without any success...

Now with a Normal distribution 1 standard deviation (SD) about for 68.3% from the mean, 2SD=95.4% and 3SD=99.7%, is this the same for a Poisson distribution and/or other parametric distributions. Also, as the mean increases in the Poisson distribution its shape changes so is the above SD% the same for all  Poisson distribution shapes OR just when it becomes sufficient to be 'Normal' like i.e mean = > ~15?

Thanks for any help, if you could include a link/source as well that would be great!

Standard deviation is different for the various distributions out there. In the case of a Poisson distribution it’s sqrt(mean) as I recall. That formula will work at all scales but, if the actual Gaussian distributed data on which your calculations are based is skewed or has some amount of kurtosis then all bets are off of course.

This looks like a reasonable link:

https://www.dur.ac.uk/physics/students/labs/skills/data/poisson/
 

If you can describe what you’re doing, then I might be able to provide some better pointers. E.g. if you’re looking at arrival rates for a queue of some type, then there any many references specific to that.

 

Link to post
Share on other sites
1 hour ago, TheBlueCat said:

Standard deviation is different for the various distributions out there. In the case of a Poisson distribution it’s sqrt(mean) as I recall. That formula will work at all scales but, if the actual Gaussian distributed data on which your calculations are based is skewed or has some amount of kurtosis then all bets are off of course.

This looks like a reasonable link:

https://www.dur.ac.uk/physics/students/labs/skills/data/poisson/
 

If you can describe what you’re doing, then I might be able to provide some better pointers. E.g. if you’re looking at arrival rates for a queue of some type, then there any many references specific to that.

 

I remember someone from the NAO blowing my mind explaining that if the rate of people joining a queue was the same as those leaving the queue, the queue would subtend to infinity. Still doesn't make sense now. 

Link to post
Share on other sites
2 minutes ago, The Generation Game said:

I remember someone from the NAO blowing my mind explaining that if the rate of people joining a queue was the same as those leaving the queue, the queue would subtend to infinity. Still doesn't make sense now. 

Well, purely mathematically, that's true:

https://www.csus.edu/indiv/b/blakeh/mgmt/documents/opm101supplc.pdf

In a single server queue, If Lambda (arrival rate) is >= to Mu (service rate), the queue length does come out of the equations as infinite. Practically, not sure it proves very much! 

Link to post
Share on other sites
4 hours ago, The XYY Man said:

Standard deviation is for poofs...

 

XYY

I think you are getting your deviations confused....just don't go shopping in M&S looking for that 'special something'  otherwise I think you are going to be disappointed! :-)

Link to post
Share on other sites
5 hours ago, TheBlueCat said:

Standard deviation is different for the various distributions out there. In the case of a Poisson distribution it’s sqrt(mean) as I recall. That formula will work at all scales but, if the actual Gaussian distributed data on which your calculations are based is skewed or has some amount of kurtosis then all bets are off of course.

This looks like a reasonable link:

https://www.dur.ac.uk/physics/students/labs/skills/data/poisson/
 

If you can describe what you’re doing, then I might be able to provide some better pointers. E.g. if you’re looking at arrival rates for a queue of some type, then there any many references specific to that.

 

Thanks @TheBlueCat...wasnt trying tocalculate anything specific, just explaining to someone, stated that all parametric (not just Normal distributions) statistics had SD and wanted to be sure. Also got me thinking if one Standard Deviation for all distribution models equates to 68.3% of the dataset, but still unable to find out if that the case.

Link to post
Share on other sites
6 minutes ago, MrXxxx said:

I think you are getting your deviations confused....just don't go shopping in M&S looking for that 'special something'  otherwise I think you are going to be disappointed! :-)

M&S...? Fuck off...!!!

Don't get nowt on me Co-op deviation card in there....

;)

 

XYY

Link to post
Share on other sites
14 minutes ago, MrXxxx said:

From what I've been told (although I never saw the pictures fortunately) the arse is pretty accurate though! :-)))

Difficult to tell from that angle, but I'm sure my arse has more skid marks than that...!

And by the way, you didn't see the pictures (plural) because there was only one picture (singular) ever posted.

Carlsberg don't do an infamous arse picture, but if they did...

 

XYY

 

Edited by The XYY Man
Link to post
Share on other sites
9 minutes ago, The XYY Man said:

Difficult to tell from that angle, but I'm sure my arse has more skid marks than that...!

And by the way, you didn't see the pictures (plural) because there was only one picture (singular) ever posted.

Carlsberg don't do an infamous arse picture, but if they did...

 

XYY

 

Well Carlsberg don't but here are two you may like to consider..:-)

image.jpeg.92053d01e7f2b03a7dfde37e393d923c.jpeg

Big Rig Releases 2nd Annual Iron Arse Ale in Support of Ride For Dad |  Canadian Beer News

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...