"the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivaria"

[zorank]

"the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivaria"

Which one is right, and why?
Zoran

 

[Rover_KE]

Re: "the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivar

It depends whether you're referring to a specific uncorrelated multivariate Poisson distribution, or any uncorrelated multivariate Poisson distribution.

We have not enough information to know, as you haven't even given us a full sentence.

(Does anybody know what the heck 'uncorrelated multivariate Poisson distribution' means? Thank goodness for copy and paste.)

Rover

 

[SoothingDave]

Re: "the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivar

Does anybody know what the heck 'uncorrelated multivariate Poisson distribution' means?

It's something to do with fish...

 

[bhaisahab]

Re: "the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivar

"Poisson" is French for fish, but what the text means and why it should mix French and English I can't imagine.

 

[SoothingDave]

Re: "the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivar

Poisson distribution | Define Poisson distribution at Dictionary.com

In all seriousness, it's a form of statistical distribution.

"Multivariate" would mean they are looking at multiple variables and "uncorrelated" would mean there is no correlation among the variables.

 

[zorank]

Re: "the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivar

It depends whether you're referring to a specific uncorrelated multivariate Poisson distribution, or any uncorrelated multivariate Poisson distribution.
Rover

Many thanks for trying to help me out.

It is actually simple. A distribution is a function. A function is a procedure for computing something from something else. In this particualr case you feed in intereger values to such function and get out values from zero to one (probabilities). Multivariate function is a product of such functions. The real problem is elswehere. It is about whether to treat a set of functions as countable or uncountable.

Sorry being so mathy here. I could have tried to construct non-math example that addresses the same issue but the flavor of it would be lost. Please bear with me. It is about the following:

I never understood this part, I have to be honest, whether "distribution" is countable or uncountable. A distribution, being a function, or an object if you wish, is parameterized by parameters. They can be real numbers (uncountable), or integers (countable).

My understanding is that an English speaking person would see the Poission distribution, which is parameterized by one real number, as an uncountable object (like sugar, water, etc). Real numbers cannot be counted, so I would expect that the set/class of all Poission distributions is uncountable. It should be either with the or without it. But then in academic writing one has to use "the" (the difinite article in front of a proper noun rule I think?).

Thus Saying "the Poisson distribution" is common. I wonder whether one can say "a Poisson distribution"? I wonder whether adding adjectives in front changes also what one has to use e.g. "multivariate Poisson distribution".

Please bear with me as this issue is a recurring pest in my writing.
All the help highly appreciated!

 

[Rover_KE]

Re: "the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivar

:sleeping:

 

[5jj]

Re: "the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivar

It is actually simple. A distribution is a function. A function is a procedure for computing something from something else. In this particualr case you feed in intereger values to such function and get out values from zero to one (probabilities). A Multivariate function is a product of such functions. The real problem is elswehere. It is about whether to treat a set of functions as countable or uncountable.

You have answered your own question here, and below.

I never understood this part, I have to be honest, whether "distribution" is countable or uncountable. A distribution, being a function, or an object if you wish, is parameterized by parameters. They can be real numbers (uncountable), or integers (countable).

My understanding is that an English speaking person would see the Poission distribution, which is parameterized by one real number, as an uncountable object (like sugar, water, etc). Real numbers cannot be counted, so I would expect that the set/class of all Poission distributions is uncountable. It should be either with the or without it.

No. Poisson distribution, like any distribution, as a way of distributing, may be uncountable, but of you have a set of Poisson distributions, then it is coutable. Real numbers can be counted, surely?

But then in academic writing one has to use "the" (the definite article in front of a proper noun rule I think?).

No. Boyle's Law, Fermat's Last Theorem, Einstein's Theory of Relativity.

Thus Saying "the Poisson distribution" is common. I wonder whether one can say "a Poisson distribution"? I wonder whether adding adjectives in front changes also what one has to use e.g. "multivariate Poisson distribution".

Well, presumably if you can have a multivariate Poisson distribution, you can have a non-multivariate one.

Your being mathy(?) is simply confusing the issue. You are dealing with language here.

 

[birdeen's call]

Re: "the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivar

Real numbers can be counted, surely?

I can't count them. I hear Chuck Norris is in the business of counting elements of infinite sets.

On a more serious note, real numbers form an uncountable set which means that there are "more" (in a set-theoretic sense) real numbers than natural numbers.

"Poisson" is French for fish, but what the text means and why it should mix French and English I can't imagine.

"Poisson" may mean fish in French, but Siméon-Denis Poisson wasn't a fish.

I have to be honest, whether "distribution" is countable or uncountable.

Probability distributions are countable in the grammatical sense. Real numbers' being uncountable and natural numbers' being countable in the set-theoretic sense has nothing to do with it. Real numbers are countable grammatically, which is proven by the fact that this sentence is correct.

I'm not suspicious usually but I'm starting to suspect your questions aren't honest.

 

[5jj]

Re: "the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivar

I'm not suspicious usually but I'm starting to suspect your questions aren't honest.

BC!

I'm shocked!!:shock:

 

[birdeen's call]

Re: "the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivar

I wonder whether one can say "a Poisson distribution"?

Yes.

 

[SoothingDave]

Re: "the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivar

Well, presumably if you can have a multivariate Poisson distribution, you can have a non-multivariate one.

That would be univariate or monovariate, wouldn't it?

 

[Tdol]

Re: "the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivar

Yes.

There are 131,000 examples of "a Poisson distribution" in Google Scholar.

Zorank, I think Google Scholar could be a simple way to test the acceptability of technical phrases. It may not be perfect, but the sources are sound, and if you get a result like that, then it's clearly in use in academic and technical writing.

 

[zorank]

Re: "the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivar

I'm not suspicious usually but I'm starting to suspect your questions aren't honest.

My questions are honest. Why wouldn't they be? (I understand that you had problems with multiple identity users etc but, once and for all, I AM NOT ONE OF THESE FOLKS!!!!!!! I AM HONESLTY FISHING FOR HELP!!!!! Why would I put my picture on and have full contact details in this forum.)

Sooooo, I will nuke the next person who accues me of being "suspicious".

I know the difference between "a apple" and "the apple" but I do not know the difference between "a Poisson distribution" and "the Poisson distribution". Why is this so hard to understand? Jesus...

 

[zorank]

Re: "the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivar

There are 131,000 examples of "a Poisson distribution" in Google Scholar.

Zorank, I think Google Scholar could be a simple way to test the acceptability of technical phrases. It may not be perfect, but the sources are sound, and if you get a result like that, then it's clearly in use in academic and technical writing.

I know. I am using it all the time and, frankly, I am getting tired of it. I realized I should spend the time to learn the principles instead. This is why I am always asking "why".

So, can someone please explain to me what is the difference in meaning between

"the Poisson distribution"

"a Poisson distribution"

?

My naive understanding is that "the Poisson distribution" = "Poisson's distribution" if the right hand side of the identity is possible. That is if I say "the Poisson distribution" I mean it as a concept, as a keyword. BUT, I am not sure.

 

[zorank]

Re: "the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivar

Probability distributions are countable in the grammatical sense. Real numbers' being uncountable and natural numbers' being countable in the set-theoretic sense has nothing to do with it. Real numbers are countable grammatically, which is proven by the fact that this sentence is correct.

Aha! That probably explains why I am so confused with this.

In a way, I presume I could say

"a real number"

right? If this is correct, then I could also say

"My favorite real numbers are pi, e, and ln 2. The numbers have to be represented with infinite number of digits."

Would that be all right to say?

 

[SoothingDave]

Re: "the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivar

It's OK to say. But those are irrational numbers you speak of. They have an infinite number of decimal digits.

2.5 and 3.0 are real numbers, but they can be expressed perfectly with a finite number of digits.

 

[Tdol]

Re: "the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivar

To me, the Poisson distribution would indicate the only possible distribution, while a Poisson distribution suggests that it is not the only possible way of viewing things. However, I should stress that I am not a statistician. In language terms, the distinction is fairly simple, but in mathematical terms, there may be a huge difference, and not one I can possibly comment on. There may be people here who are qualified to discuss the mathematics, but they will be a tiny minority (like SoothingDave). I certainly cannot answer these questions with any degree of certainty as they belong to an area I know little about.

I honestly think you would be better off looking for the answers in an English-language advanced mathematics forum. My wife's an epidemiologist who does use such maths, but I still wouldn't be able to go through her to answer your questions. They are, at heart, mathematical rather than linguistic.

 

[zorank]

Re: "the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivar

...an English-language ... mathematics forum...

I've spent ages looking for it and, alas, haven't found it yet :-(. You guys are closest I could come up with :-D. Of course, if there is indeed such a forum, I would be happy for a reference/link (and you would get ridd of me ;-)). In the meantime, I will try to tone the math down (or try with an equivalent math free example). But please, let me go on with this particular example for a little while...

Regards
Zoran

 

[zorank]

Re: "the uncorrelated multivariate Poisson distribution" or "an uncorrelated multivar

It's OK to say. But those are irrational numbers you speak of. They have an infinite number of decimal digits.

2.5 and 3.0 are real numbers, but they can be expressed perfectly with a finite number of digits.

Fine. I am not into numbers really. What about distributions that are labeled (parametereized) by real numbers.

For example, there is this sentence from wikipedia:

"In probability theory and statistics, the Poisson distribution ... is a discrete probability distribution that expresses the probability of a...."

Why do they use the? Would it be all right to say

"In probability theory and statistics, a Poisson distribution ... is a discrete probability distribution that expresses the probability of a...."