Is it okay to post just identities on arXiv?

[mesa]

If I have some new identities that are very powerful but don't have time to go into the details is it okay to simply post just the functions on arXiv for now?

A quick check on wolfram as to the validity of the functions will confirm all my work.

 

[jesse73]

No.

The details are what is important in physics and if you can check the validity of it in mathematica then it is probably known anyhow.

 

[dipole]

Agreed, if you just want to share something useful, post it on a blog or even in the math forum. Scientific journals (even on-line ones) require a much higher standard.

 

[mesa]

No.

The details are what is important in physics and if you can check the validity of it in mathematica then it is probably known anyhow.

Three Professor's on campus said they have not seen it before, but that certainly does not mean it isn't known. I freed up the remainder of the day to spend in the math department to see what else I can dig up.

Agreed, if you just want to share something useful, post it on a blog or even in the math forum. Scientific journals (even on-line ones) require a much higher standard.

I'm not surprised. Thank you both for the input.
If either of you have suggestions as to what is the best way to find out information like this then feel free to share. The identities are infinite series summations for the gamma function, and all real numbers. Both also work for all imaginary values as well.

 

[mesa]

Update, after a long conversation with some faculty they are sending me upstairs to speak with the research department since they have never seen these identities before, how exciting!

Wish me luck!

 

[Ben Niehoff]

If you have something original, then it deserves at least a short write-up. There ought to be a math journal that focuses mainly on "letters", which are papers in the 4-10 page range (even 10 might be pushing it).

In any case, simply listing a bunch of identities is not helpful, since if they are truly original, it will not be clear to the reader whether they are true or not (for example, Ramanujan is famous for writing things down without proof; and while some of it is brilliant, some of it is also wrong!). A write-up where you explain how to obtain the identities, or how to show they are true, will be much more helpful to the reader.

 

[dipole]

I would consult a book on mathematical functions to check if it's been published - the Gamma function is very well studied, so it's likely what you have is known. This might be a good place to start too: http://dlmf.nist.gov/5

However, if you can't find your representation anywhere, it'd definitely be worth publishing - proofs can often be made easier by just using a specific form a function which you're using, so someone would probably find it useful at some point.

 

[mesa]

If you have something original, then it deserves at least a short write-up. There ought to be a math journal that focuses mainly on "letters", which are papers in the 4-10 page range (even 10 might be pushing it).

In any case, simply listing a bunch of identities is not helpful, since if they are truly original, it will not be clear to the reader whether they are true or not (for example, Ramanujan is famous for writing things down without proof; and while some of it is brilliant, some of it is also wrong!). A write-up where you explain how to obtain the identities, or how to show they are true, will be much more helpful to the reader.

Okay, the proof is straightforward for the first identity and going from there to the others is easy to do. I don't think it would take more than 2 to 3 pages at the most to write it, but then again I have never done this before.

I would consult a book on mathematical functions to check if it's been published - the Gamma function is very well studied, so it's likely what you have is known. This might be a good place to start too: http://dlmf.nist.gov/5

However, if you can't find your representation anywhere, it'd definitely be worth publishing - proofs can often be made easier by just using a specific form a function which you're using, so someone would probably find it useful at some point.

Good news, it is not in your link! I have also checked the standards as well, wikipedia, wolframalpha, combined with strenuous google searches. I am new at this so if you have other recommendations on where else I should look I am open to it.

One of the Professor's I spoke with was astonished at how quickly this Gamma function converges and said it would be excellent for computations, perhaps I should be looking at computer science as well to look for functions that are similar (or the same)?

Either way, it went very well today, I met with 3 research Professors and have a 4th to visit shortly. None of them had seen it before but math is huge so it doesn't mean it is 'new' but I remain hopeful!

 

[dipole]

I can't give you much advice about looking into the literature because I'm not a mathematician and not particularly familiar with math journals. If it converges quickly, that's another reason to publish it. You should ask the next professor about advice on looking into this.

If you do decide to publish it in a peer-review journal, you should talk to the department head and see if they can pay to have it submitted - normally one must pay a fee to submit a paper for review.

 

[mesa]

I can't give you much advice about looking into the literature because I'm not a mathematician and not particularly familiar with math journals. If it converges quickly, that's another reason to publish it. You should ask the next professor about advice on looking into this.

If you do decide to publish it in a peer-review journal, you should talk to the department head and see if they can pay to have it submitted - normally one must pay a fee to submit a paper for review.

Shouldn't I start with arXiv (especially considering this being my first), post a link here, have the PF community beat me up to hone it into something decent and then look for publishing?

 

[Ben Niehoff]

Wait, are you saying you have fast-converging infinite series that sum to the Gamma function, and Mathematica is able to verify your expressions?

If Mathematica is able to do it, then it is probably already known. Check Abramowitz and Stegun. There is a lot of obscure stuff in there that no reasonable person can keep in their working memory (i.e., your math professors might think you have something original even if it's quite well known).

Also, now you're also claiming your series converge on the whole real line, which is nonsense, because the Gamma function has poles...

Anyway, if after checking A&S (and possibly some other sources) you are still convinced you have something original, then put it on arXiv and let the mathematical community critique it. PF is not a peer review site.

 

[mesa]

Wait, are you saying you have fast-converging infinite series that sum to the Gamma function, and Mathematica is able to verify your expressions?

If Mathematica is able to do it, then it is probably already known. Check Abramowitz and Stegun. There is a lot of obscure stuff in there that no reasonable person can keep in their working memory (i.e., your math professors might think you have something original even if it's quite well known).

I don't see it, how wonderful! If it is out there then why wouldn't the simplest fastest converging series (from what we have seen so far) be well known? Not saying this isn't the case but that would be odd would it not?

Apparently there is a Professor helping with the search (just sent me an email), he discovered a 'similar' function for taking partial derivatives (a subject admittedly above my head) but it is not used for gamma, it just happens to be the closest thing so far.

Also, now you're also claiming your series converge on the whole real line, which is nonsense, because the Gamma function has poles...

I should specify the derived function is 1/gamma. Taking the inverse of both sides gives gamma. Thank you for the correction!

Anyway, if after checking A&S (and possibly some other sources) you are still convinced you have something original, then put it on arXiv and let the mathematical community critique it. PF is not a peer review site.

I am aware that PF is not peer review especially considering we do not allow for 'personal theories' as per guidelines although I respect the opinion of the community and as such will encourage input after posting on arXiv.

Very good information and link Ben!

 

[1MileCrash]

This must be exciting for you, whether it is known or not, you should be proud because you still found it on your own.

 

[mesa]

This must be exciting for you, whether it is known or not, you should be proud because you still found it on your own.

Absolutely, it's a win win!

I can't wait to see what is to come, each semester there are more tools to play with. I already have a stack of notebooks a foot and a half deep of just personal exploration and the pile thickens faster with the passing of each new course. We live in a wonderful time!

 

[robphy]

I should specify the derived function is 1/gamma. Taking the inverse of both sides gives gamma. Thank you for the correction!

So, I presume you have looked at
http://www.google.com/search?q=reciprocal+gamma
http://www.google.com/search?q=factorielle+weierstrass

In addition to the identities themselves,
there might be publishable interest in the method you use to obtain them (and possibly other similar identities) and how this identity might be helpful in solving certain problems.

(I have found some curious identities [on something completely different] as well... but haven't found good contexts in which to write them up. They are on the backburner, while I work on other more pressing things.)

 

[mesa]

So, I presume you have looked at
http://www.google.com/search?q=reciprocal+gamma
http://www.google.com/search?q=factorielle+weierstrass

In addition to the identities themselves,
there might be publishable interest in the method you use to obtain them (and possibly other similar identities) and how this identity might be helpful in solving certain problems.

(I have found some curious identities [on something completely different] as well... but haven't found good contexts in which to write them up. They are on the backburner, while I work on other more pressing things.)

Yup, google has turned up nothing like it.

The method for deriving it has unfortunately drawn a few rolled eyes but I have a solid proof based off of another well known (and proven) series so was thinking to just leave it at that but this is causing issues as well (many Professors are complaining it is too simple, arrrggghhh! I thought that was the whole point?!?)

What kind of identities do you have? If you have anything like these maybe we should share and see if we can gain greater insight.

Here are are a few on my list (the first two are general solutions that work for all real and imaginary values),

$$ \frac{1}{\Gamma(k)}=\sum_{n=0}^{\infty} f(n,k)$$
$$\frac{m}{k}=\sum_{n=0}^{\infty} f(n,m,k)$$

For the $$\frac{m}{k}=\sum_{n=0}^{\infty}...$$ I now have two completely different working functions which should lead to a second general solution for 1/gamma(k). I am getting close on a new set of trig identities but there is still much work to be done.

I also have a whole bunch of standard series identities for 'e', a couple for 'i', ln (2), Pi, etc. and one I just finished today that will generate an infinite number of 'unique' infinite series summations exactly equal to the square root of Pi.

Let me know!

 

[robphy]

One set of identities involve the Einstein tensor. Another set involves determinants of sums of matrices.
So, these are quite different from your identities.

But keep working on it... and keep an eye out for what is going on in the literature (current and past).
You might get a sense of where you might try to publish... and what their standards of publication are.

You also might consider giving a talk on your ideas in your department.

 

[mesa]

One set of identities involve the Einstein tensor.

I do not have (yet) the mathematics (or physics) for those types of things, it must be wonderful to know these areas well enough to produce identities!

Another set involves determinants of sums of matrices.

Now there is something I at least understand the fundamentals of...

So, these are quite different from your identities.

I would say so
So why haven't you contributed your work yet?

But keep working on it... and keep an eye out for what is going on in the literature (current and past).
You might get a sense of where you might try to publish... and what their standards of publication are.

You also might consider giving a talk on your ideas in your department.

The little literature I have read seems 'complex' with contributors spending little time trying to re-imagine the fundamentals (which is worrisome).

Are papers rejected if they don't fit into the mainstream of current mathematics even if it does present something new? Or is it that most people just don't play with such things anymore and instead lean towards higher orders of abstraction and complexity?

I will continue to read but I am really enjoying my time just knocking on doors and getting one on one time with our Professors. Some are more welcoming than others but I usually have my most interesting conversations with those who keep their doors mostly (but not all the way) closed.