Dtjytyk

Is there a minimum number of transactions in a block?

How does one intimidate enemies without having the capacity for violence?

What makes Graph invariants so useful/important?

Copenhagen passport control - US citizen

Should I join an office cleaning event for free?

How can the DM most effectively choose 1 out of an odd number of players to be targeted by an attack or effect?

What do you call something that goes against the spirit of the law, but is legal when interpreting the law to the letter?

Why is an old chain unsafe?

declaring a variable twice in IIFE

What Brexit solution does the DUP want?

Is it possible to do 50 km distance without any previous training?

Why is the design of haulage companies so “special”?

Can you lasso down a wizard who is using the Levitate spell?

Japan - Plan around max visa duration

Example of a relative pronoun

A function which translates a sentence to title-case

I probably found a bug with the sudo apt install function

Are tax years 2016 & 2017 back taxes deductible for tax year 2018?

Draw simple lines in Inkscape

What is the offset in a seaplane's hull?

What is the white spray-pattern residue inside these Falcon Heavy nozzles?

I see my dog run

Calculus Optimization - Point on graph closest to given point

Can Medicine checks be used, with decent rolls, to completely mitigate the risk of death from ongoing damage?

Are events correlated in a Poisson distribution?

Correlated Poisson Distributionpoisson distributionPoisson Distribution - Poor UnderstandingPoisson Distribution: What's the probability of getting a first week without any events when you are told that 5 events occurred within a month?Poisson DistributionPoisson distribution CDFProbability of overlapping events in Poisson distributionPoisson distribution, and embedded poisson distributionInverse poisson distribution calculationPoisson distribution probability (another)

0

$begingroup$

I would like some input on a discussion I am currently having with a colleague regarding some measurements that were done recently.

I have some radiation physics data that is processed as a histogram. This histogram I believe can be modeled as a Poisson distribution since each value that goes into a given interval, or bin, is independent of any other. My colleague is arguing that this thinking is incorrect, but rather, bins that are closer together in the histogram are correlated, and as a result I am over-estimating the amount of error in my measurements.

Is this correct? To me, the formulation of a histogram is simply an intuitive data processing technique, wherein there is no "cross-talk" between bins.

There is obviously quite a bit of overlap here with regard to physics, so I will resort to analogy for the lay person: We are breaking-up heavy ions using shielding of varying types. We could refer to the heavy ion fragments as pieces of "mail", the detector as the "mailbox", a histogram bin as the "delivery day", and the total number of bins containing all the mail as the "month". Of course, integration over the histogram bins gives the total amount of mail delivered in a month. The peak delivery day in a given month is dependent on the amount of shielding between the beam and the detector. Following this analogy, my colleague is arguing that any two delivery days are correlated.

statistics poisson-distribution

share|cite|improve this question

edited Mar 20 at 14:46

Joel DeWitt

asked Mar 20 at 13:06

Joel DeWittJoel DeWitt

1012

$endgroup$

• 
  
  
  

  
  1
  

  
  
  
  
  
  

  
  $begingroup$
  I think you are right. Radioactive decay is the classic Poisson process. The histogram is just a data visualization tool.
  $endgroup$
  – Ethan Bolker
  Mar 20 at 13:09
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  $begingroup$
  I would say that it depends on the underlying processes. If your observation is a superposition of two effects you would have a problem in using the histogram. E.g. some nuclear active atoms tend to decay in a special sequence. So it primarily depends on the underlying physics.
  $endgroup$
  – MachineLearner
  Mar 20 at 13:13
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  $begingroup$
  @JoelDeWitt Not an expert in this field :D. But if an ion breaks with more intensity can it be reflected by the shielding? That would intensify the frequency of breaking ... (interpret everything that I say from the perspective of a child that does not know anything about physics #NotAPhysisct :D). I am thinking about the double slit experiment, which also influences the distribution of particles on the shield. Also if something breaks off will it not be less likely to break something off again from the same atom because the excitation is reduced?
  $endgroup$
  – MachineLearner
  Mar 20 at 13:28
  
  
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  $begingroup$
  @MachineLearner, I have updated my question with an analogy to make it easier to understand.
  $endgroup$
  – Joel DeWitt
  Mar 20 at 14:47
  

  
  
  
  

add a comment | 

0

$begingroup$

I would like some input on a discussion I am currently having with a colleague regarding some measurements that were done recently.

I have some radiation physics data that is processed as a histogram. This histogram I believe can be modeled as a Poisson distribution since each value that goes into a given interval, or bin, is independent of any other. My colleague is arguing that this thinking is incorrect, but rather, bins that are closer together in the histogram are correlated, and as a result I am over-estimating the amount of error in my measurements.

Is this correct? To me, the formulation of a histogram is simply an intuitive data processing technique, wherein there is no "cross-talk" between bins.

There is obviously quite a bit of overlap here with regard to physics, so I will resort to analogy for the lay person: We are breaking-up heavy ions using shielding of varying types. We could refer to the heavy ion fragments as pieces of "mail", the detector as the "mailbox", a histogram bin as the "delivery day", and the total number of bins containing all the mail as the "month". Of course, integration over the histogram bins gives the total amount of mail delivered in a month. The peak delivery day in a given month is dependent on the amount of shielding between the beam and the detector. Following this analogy, my colleague is arguing that any two delivery days are correlated.

statistics poisson-distribution

share|cite|improve this question

edited Mar 20 at 14:46

Joel DeWitt

asked Mar 20 at 13:06

Joel DeWittJoel DeWitt

1012

$endgroup$

• 
  
  
  

  
  1
  

  
  
  
  
  
  

  
  $begingroup$
  I think you are right. Radioactive decay is the classic Poisson process. The histogram is just a data visualization tool.
  $endgroup$
  – Ethan Bolker
  Mar 20 at 13:09
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  $begingroup$
  I would say that it depends on the underlying processes. If your observation is a superposition of two effects you would have a problem in using the histogram. E.g. some nuclear active atoms tend to decay in a special sequence. So it primarily depends on the underlying physics.
  $endgroup$
  – MachineLearner
  Mar 20 at 13:13
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  $begingroup$
  @JoelDeWitt Not an expert in this field :D. But if an ion breaks with more intensity can it be reflected by the shielding? That would intensify the frequency of breaking ... (interpret everything that I say from the perspective of a child that does not know anything about physics #NotAPhysisct :D). I am thinking about the double slit experiment, which also influences the distribution of particles on the shield. Also if something breaks off will it not be less likely to break something off again from the same atom because the excitation is reduced?
  $endgroup$
  – MachineLearner
  Mar 20 at 13:28
  
  
  

  
  
  
  


• 
  
  
  

  
  

  
  
  
  
  
  

  
  $begingroup$
  @MachineLearner, I have updated my question with an analogy to make it easier to understand.
  $endgroup$
  – Joel DeWitt
  Mar 20 at 14:47
  

  
  
  
  

add a comment | 

$begingroup$

I would like some input on a discussion I am currently having with a colleague regarding some measurements that were done recently.

I have some radiation physics data that is processed as a histogram. This histogram I believe can be modeled as a Poisson distribution since each value that goes into a given interval, or bin, is independent of any other. My colleague is arguing that this thinking is incorrect, but rather, bins that are closer together in the histogram are correlated, and as a result I am over-estimating the amount of error in my measurements.

Is this correct? To me, the formulation of a histogram is simply an intuitive data processing technique, wherein there is no "cross-talk" between bins.

There is obviously quite a bit of overlap here with regard to physics, so I will resort to analogy for the lay person: We are breaking-up heavy ions using shielding of varying types. We could refer to the heavy ion fragments as pieces of "mail", the detector as the "mailbox", a histogram bin as the "delivery day", and the total number of bins containing all the mail as the "month". Of course, integration over the histogram bins gives the total amount of mail delivered in a month. The peak delivery day in a given month is dependent on the amount of shielding between the beam and the detector. Following this analogy, my colleague is arguing that any two delivery days are correlated.

statistics poisson-distribution

share|cite|improve this question

edited Mar 20 at 14:46

Joel DeWitt

asked Mar 20 at 13:06

Joel DeWittJoel DeWitt

1012

$endgroup$

share|cite|improve this question

edited Mar 20 at 14:46

Joel DeWitt

asked Mar 20 at 13:06

Joel DeWittJoel DeWitt

1012

share|cite|improve this question

edited Mar 20 at 14:46

Joel DeWitt

asked Mar 20 at 13:06

Joel DeWittJoel DeWitt

1012

asked Mar 20 at 13:06

Joel DeWittJoel DeWitt

1012

asked Mar 20 at 13:06

Joel DeWittJoel DeWitt

1012