Arp objects, QSOs, Statistics

[BeAChooser]

The difference is that I created a set of "Amaik peaks", from certain inputs.

What inputs?

And could you tell us the *inputs* you used to derive the DRDS peaks?

The "regulars" are another test of this same thing - they are derived by simply dividing up the range [0, 3] to produce 7 regularly-spaced "peaks" (oh, and I expressed all of them as having only two significant digits, so for example 1.13 rather than 1.125).

Ok, that I understand now.

Now here are some questions about the source of your seven data sets. How did you "sample" them? I think it's fairly clear that Arp et al, got their cases by taking every instance of a large number of quasars near a galaxy they could find. Now obviously, you have access to newer and better data than they did at the time they first identified their cases. Did you provide us with all of them you have identified? Or did you select them from a larger database in some manner? If so, how? How did you make sure they were random and not biased?

Also, you indicated that "at least one set is 'mock' QSOs". Which case(s) are those? And how did you go about creating these "mock" QSO redshifts? How did you go about ensuring that you weren't biasing them in some manner?

Also, you indicated one case is for a group of quasars centered on a random point in the sky. How did you pick that random point. And did you actually look to see if there are galaxies nearby? Can to provide us with all the information you have about this case? If not, why not?

Also, in some of the cases you said the quasars were within 20 degrees of a random direction. Precisely, what do you mean by that?

 

[Dancing David]

What inputs?

And could you tell us the *inputs* you used to derive the DRDS peaks?



Ok, that I understand now.

Now here are some questions about the source of your seven data sets. How did you "sample" them? I think it's fairly clear that Arp et al, got their cases by taking every instance of a large number of quasars near a galaxy they could find. Now obviously, you have access to newer and better data than they did at the time they first identified their cases. Did you provide us with all of them you have identified? Or did you select them from a larger database in some manner? If so, how? How did you make sure they were random and not biased?

Also, you indicated that "at least one set is 'mock' QSOs". Which case(s) are those? And how did you go about creating these "mock" QSO redshifts? How did you go about ensuring that you weren't biasing them in some manner?

Also, you indicated one case is for a group of quasars centered on a random point in the sky. How did you pick that random point. And did you actually look to see if there are galaxies nearby? Can to provide us with all the information you have about this case? If not, why not?

Also, in some of the cases you said the quasars were within 20 degrees of a random direction. Precisely, what do you mean by that?

Gee, it is sort of ironic, all these questions from someone who refuses to answer direct questions.

DRD is very open as opposed to you and your selective answers.

 

[Reality Check]

The "disease" in this case is an assocation of a large number of quasars with a galaxy and/or features of the galaxy. That's how they identified it. Do you have any evidence that Arp, et. al., didn't report every case they knew about where there were large numbers of quasars or apparent associations of large number of quasars with minor axis or other features of galaxies?

Lets see... A small number of papers spread over several years each reporting a few (or one) instance. I take this as evidence that every case has not been reported and that the papers are not reporting a sample - they are reporting a collection of examples.

Perhaps you can point the paper with Arp's entire sample collected in it and staistics that are calculated from the sample as a whole (not probabilites for a single instance).

 

[BeAChooser]

Perhaps you can point the paper with Arp's entire sample collected in it and staistics that are calculated from the sample as a whole (not probabilites for a single instance).

Wish I could but I can't. It just seems that Arp et. al. were highlighting ANY case that seemed to show an inordinate number of quasars near a galaxy or lined up with some feature of a galaxy, so I take that as evidence what they offered at the time was a complete listing of all such cases that they were aware existed. Thus, that would constitute a random sample from all such cases that actually do exist.

You might find this of interest:

http://www.haltonarp.com/articles/origins_of_quasars_and_galaxy_clusters

In order to separate background quasars from candidates for association with M101 another SIMBAD screen was set at the relatively bright apparent magnitude of V = 17.1 mag. The search was supplemented by visual search of the Veron and Veron Catalog and checks with NED lists of high redshift quasars. The quasars found are plotted in Fig. 5. The first impression is that these brightest quasars are distributed along the same general line of Markarian galaxies as just discussed. In detail, there are quasars near each of the plus signs which represent the Markarian galaxies from Fig. 4, suggesting that these quasars may have originated more recently from these lower redshift, active Markarian galaxies and not necessarily as direct ejections from M101 itself.

Figure 5
http://www.haltonarp.com/articles/origins_of_quasars_and_galaxy_clusters/illustrations/figure_5.png
CAPTION: The plus signs are the same Markarian galaxies from Fig. 4 with now all QSO's less than z = 17.1 mag. added as filled circles. All known 3C radio objects in the area are represented by open circles. The two circled plus signs are Abell Clusters from Fig. 1 .

Some of the high redshift quasars here have such bright apparent magnitudes that whatever the average luminosity for this redshift may be, they are certainly among the closest of this class - if the red shift does not indicate distance (Arp 1999). Examples are redshifts of z = 2.63 at 17.0 mag., z = 1.86 at 16.6 mag. and z = 3.19 at 15.8 mag. (We should remark that while the z = 3.19 quasar is not near a Markarian galaxy it is quite near an infrared, IRAS, galaxy of 15.3 mag. and z = .037 and therefore similar to the Markarian galaxies plotted here.) The 15.5 mag. object is a BL Lac type quasar, OQ 530, very bright in apparent magnitude for its class and agreeing with the close association of BL Lac's with nearby galaxies as reported in Arp (1997, Table 2 and 1998a).

Numerical Coincidences of Redshifts

There are some remarkable numerical agreements in redshifts of a number of objects in Fig. 5. The quasars at z = .646, .660 and .656 stand out. It should be remarked that there is an additional quasar at 17.7 mag., slightly below the cut off of 17.1 mag., which has a z = .646 and falls just SW of M101. The latter makes a close apparent pair with the z = .660 quasar across M101 directly.

... snip ...

The numerical coincidence of the 3C quasars at z = .961 and .967 also stands out. Even more strikingly, these redshifts agree very closely with peak values of redshifts in the Karlsson formula which expresses the empirical relation found for many years for quasar redshifts:

z = .06; .30; .60; .96; 1.41; 1.96; 2.64; 3.48...

 

[Reality Check]

Periodicities of Quasar Redshifts in Large Area Surveys (H. Arp, C. Fulton, D. Roscoe) does mention their methodolgy

We should emphasize that the procedure we are following is to examine apparent groups and concentrations of quasars that appear to be physically associated. In each of the cases investigated here it turns out there is a brighter, active galaxy present which is a candidate for the origin of these quasars. We then proceed to test this identification by seeing whether the disparate redshifts are brought into the order of the Karlsson formula when their redshifts are transformed to this chosen parent.

but once again they seem to be only presenting a subset of their cases, i.e. only those with a brighter, active galaxy present. This leads to lots of probably naive questions:
What about the cases with 2 or more brighter, active galaxies present?
Why does the galaxy have to be bright?
What is the criteria for brightness?

I am also unsure about why the Karlsson formula should identify the quasars as associated with the bright galaxy. If they are assuming that the Karlsson formula is caused by the ejection of quasars from bright galaxies then they cannot use it to test that quasars are ejected from bright galaxies. That looks like circular reasoning.

On the other hand if the Karlsson formula is caused by something else (e.g. a dependence of quasar formation on the large scale structure of the universe) then they still cannot use it.

 

[BeAChooser]

What about the cases with 2 or more brighter, active galaxies present?
Why does the galaxy have to be bright?
What is the criteria for brightness?

All of those questions are addressed in the paper that you linked. Did you not bother to read it? Apparently not, based on the rest of what you wrote. But thanks for posting these additional examples suggesting there's a problem with the mainstream assumption that redshifts are only related to distance. Perhaps I'll get around to running them through my equation and see what pops up.

 

[Dancing David]

Wish I could but I can't. It just seems that Arp et. al. were highlighting ANY case that seemed to show an inordinate number of quasars near a galaxy or lined up with some feature of a galaxy, so I take that as evidence what they offered at the time was a complete listing of all such cases that they were aware existed. Thus, that would constitute a random sample from all such cases that actually do exist.

No that is a selected sample, so whats the deal you select galaxies that have a 'high number of QSOs near them" and then you turn around and call that random?

That is called selecting BAC.

You might find this of interest:

http://www.haltonarp.com/articles/origins_of_quasars_and_galaxy_clusters

 

[Dancing David]

from the mouth of Arp:

We should emphasize that the procedure we are following is to examine apparent groups and concentrations of quasars that appear to be physically associated. In each of the cases investigated here it turns out there is a brighter, active galaxy present which is a candidate for the origin of these quasars.

So it is not a random sample, it is a selected sample, and Arp even gives the criteria for selection.

That is his sample bias that is bolded. Geesh BAC, I know you do this on purpose, you are not dense, you are quite smart. So do the poeple in your newletter fan club just lap this stuff up?

 

[Reality Check]

All of those questions are addressed in the paper that you linked. Did you not bother to read it? Apparently not, based on the rest of what you wrote. But thanks for posting these additional examples suggesting there's a problem with the mainstream assumption that redshifts are only related to distance. Perhaps I'll get around to running them through my equation and see what pops up.

What about the circular reasoning?

 

[DeiRenDopa]

The difference is that I created a set of "Amaik peaks", from certain inputs.

What inputs?

And could you tell us the *inputs* you used to derive the DRDS peaks?

The "regulars" are another test of this same thing - they are derived by simply dividing up the range [0, 3] to produce 7 regularly-spaced "peaks" (oh, and I expressed all of them as having only two significant digits, so for example 1.13 rather than 1.125).

Ok, that I understand now.

Now here are some questions about the source of your seven data sets. How did you "sample" them? I think it's fairly clear that Arp et al, got their cases by taking every instance of a large number of quasars near a galaxy they could find. Now obviously, you have access to newer and better data than they did at the time they first identified their cases. Did you provide us with all of them you have identified? Or did you select them from a larger database in some manner? If so, how? How did you make sure they were random and not biased?

Also, you indicated that "at least one set is 'mock' QSOs". Which case(s) are those? And how did you go about creating these "mock" QSO redshifts? How did you go about ensuring that you weren't biasing them in some manner?

Also, you indicated one case is for a group of quasars centered on a random point in the sky. How did you pick that random point. And did you actually look to see if there are galaxies nearby? Can to provide us with all the information you have about this case? If not, why not?

Also, in some of the cases you said the quasars were within 20 degrees of a random direction. Precisely, what do you mean by that?

Good to see that you have not put me on ignore BAC!

Of course I'd be more than happy to address your questions.

First, however, I'd really appreciate it if you could tell me whether the calculations I have performed, using the approach you took so many posts (and so much time) to develop and explain, are correct or not.

If they are, a simple 'yep, all correct' will do.

If they are not, then would you please take just one, and go through it in detail, pointing out where I went wrong?

Note that the calculations themselves have nothing to do with any interpretation of the resulting numbers - calculations are just dumb, mechanical things.

 

[DeiRenDopa]

In this post I'll be looking at the BAC calculation on its own, 'blind' to where it came from, what interpretations can (or cannot) be drawn from the 'probabilities', etc.

Recall that BAC's post #329 contains a reference to several redshifts, one of which is 2.63. Recall that in my post #496 I derived a value of zero for the 'regular peaks'.

Let's look at how this comes about.

The BAC calculation involves multiplying several numbers together; there's no addition or subtraction involved once the numbers are derived.

Most of the numbers multiplied together are 'dz's, which are calculated by subtracting one number from another and taking the absolute value.

The two numbers involved in the subtraction are a 'redshift' and a 'peak value' (as the absolute value is used, it doesn't matter which is subtracted from which).

It follows that in every case where a redshift is the same as a peak value, the output 'probability' will always be zero.

And so it is for one of the NGC 2639 input redshifts (2.63) and one of the 'regular peaks' (2.63); the output 'probability' must be zero.

Of course, it can happen for any set of 'peaks'; for example, in set FIVE one redshift is 0.963945. If this had been rounded to 2 decimal places (0.96) it would have been the same as one of the Karlsson peaks (0.96), and the output 'probability' would have been zero.

Next: now we know what is required for the output 'probability' to be zero, let's look at how high it can go; for example, can it ever be >1?

 

[DeiRenDopa]

In this post I'll continue to look at the BAC calculation on its own, 'blind' to where it came from, what interpretations can (or cannot) be drawn from the 'probabilities', etc.

Recall that the Karlsson peaks have values of 0.06, 0.3, 0.6, 0.96, 1.41, 1.96, and 2.64; 7 peaks in all (n_k = 7, in other words).

The dz's for z's that are half-way between each peak are thus 0.12, 0.15, 0.18, 0.225, 0.275, and 0.34.

Suppose we have two z's, each with values 2.3, i.e. half-way between the two highest Karlsson peaks. What is the 'BAC probability'? Here r = 2, so crunching the numbers we get 2.52 (rounding to two decimal places).

Suppose there are six such z's; the 'BAC probability' works out to be 15.96.

We get the same 'BAC probability' > 1 for any number of z's (>2), for z = 1.685 (half-way between the 2nd and 3rd highest peaks), and z = 1.185 (3rd and 4th highest peaks).

And so it goes; whenever the dz's are > 0.214 (approx), the 'BAC probability' will be >1.

Can anyone explain why?

Now if there were six z's with values of 0.18, 0.45, 0.78, 1.185, 1.685, and 2.3 - i.e. half-way between each of the seven Karlsson peaks - the 'BAC probability' would be 0.70.

Playing around with different combinations of the 'half-way' z's - two of 2.3, no 0.18, and one each of the rest, for example - shows that many yield 'BAC probabilities' > 1.

So it's quite easy to come up with 'pathological' cases where the 'BAC probability' is > 1.

Comments?

Next: how does the 'BAC probability' change as r increases, for random sets of input z's?

 

[DeiRenDopa]

In this post I'll continue to look at the BAC calculation on its own, 'blind' to where it came from, what interpretations can (or cannot) be drawn from the 'probabilities', etc.

I generated 300 random numbers in [0, 3]. The random number generator supposedly gives a uniform distribution, and the numbers in bins of 0.5 are 58, 41, 44, 53, 53, and 51 (so it seems that is so ... anyone want to do a statistical test?).

I found the nearest 'Karlsson peak' to each random number, and calculated the 'dz' for each (there are four dz's < 0.005 - what is the expected number?).

I then took sets of six, ten, and 20 dz's (295, 291, 281 sets, respectively; I was a little lazy), and calculated the 'BAC probabilities'.

They range from 1.7x10^-6 to 0.98, with an average of 0.04 (sets of six, two significant figures); 3.5x10^-8 to 0.66 (ten), with an average of 0.008; and 7.2x10^-14 to 0.001, with an average of 8.9x10^-6 (20).

So it seems the more input 'redshifts' to the 'BAC calculation', the lower the 'BAC probability'.

Comments?

Next: how do the other three distributions (Amaik, regular, and DRD) behave? Specifically, under what conditions can they give a 'BAC probability' of zero? Is it possible for them to produce 'BAC probabilities' > 1? Do the 'BAC probabilities' also get smaller as the number of input 'redshifts' increases?

 

[DeiRenDopa]

One small extra: the average of the 300 dz's is 0.124 (to 3 significant digits).

With 6/10/20 'input redshifts' of 0.124, the 'BAC probability' is 0.038, 0.0042, and 1.8 x 10^-5, respectively.

Compare these with averages in my last post.

If the input z's are truly random in [0, 3], what value should the dz's converge to?

ETA: in my previous post, I gave two numbers to only one significant figure (0.04 and 0.008); to two these should be 0.042 and 0.0079, respectively.

 

[DeiRenDopa]

... snip ...

If the input z's are truly random in [0, 3], what value should the dz's converge to?

... snip ...

That should read

... what value should the average of the dz's converge to?

 

[DeiRenDopa]

In this post I'll continue to look at the BAC calculation on its own, 'blind' to where it came from, what interpretations can (or cannot) be drawn from the 'probabilities', etc.

Recall that the Karlsson peaks have values of 0.06, 0.3, 0.6, 0.96, 1.41, 1.96, and 2.64; 7 peaks in all (n_k = 7, in other words).

The dz's for z's that are half-way between each peak are thus 0.12, 0.15, 0.18, 0.225, 0.275, and 0.34.

Suppose we have two z's, each with values 2.3, i.e. half-way between the two highest Karlsson peaks. What is the 'BAC probability'? Here r = 2, so crunching the numbers we get 2.52 (rounding to two decimal places).

Suppose there are six such z's; the 'BAC probability' works out to be 15.96.

We get the same 'BAC probability' > 1 for any number of z's (>2), for z = 1.685 (half-way between the 2nd and 3rd highest peaks), and z = 1.185 (3rd and 4th highest peaks).

And so it goes; whenever the dz's are > 0.214 (approx), the 'BAC probability' will be >1.

Can anyone explain why?

... snip ...

Comments?

... snip ...

Hmmm ...

No one want to try to explain why? And no comments?

OK ... I found a very curious thing: 3/14 = 0.2142857...

I wonder if "> 0.214 (approx)" can be replaced by "> 3/14" ?

And if so, why?

And where did this "3/14" come from anyway?!?

 

[robinson]

7 post in a row. And nobody responded. I know you are new, so let me give you a clue. If nobody is responding, let it die a quiet and dignified death.

 

[DeiRenDopa]

7 post in a row. And nobody responded. I know you are new, so let me give you a clue. If nobody is responding, let it die a quiet and dignified death.

What, and miss out on another page or five of posts I can add to the 'Questions BAC has run away from' thread (or whatever it's called)?

You of all JREF forum members should surely appreciate the value of holding people to account for what they write, shouldn't you*? I mean, especially given the venom, bombast, ridicule, and more heaped upon those who didn't go along with BAC's certainty over what we might call, in shorthand, 'the Arp Karlsson peak quasars'?

And surely if there's anything to 'Arp's statistics', given how dramatically it would change not only astronomy but also physics, don't you think it worthwhile to run it to ground (and show just how full of errors of many different kinds this idea/these ideas is/are ... or not)?

* at least in this section of the forum, and on matters with clear scientific content.

 

[Dancing David]

7 post in a row. And nobody responded. I know you are new, so let me give you a clue. If nobody is responding, let it die a quiet and dignified death.

No, it is just that i am not going to try to understand the bogus math that BAC uses, I think that what DRD is doing just proves that BAC is making it up as they go along.

As of yet BAC has refused to address the issue of the thread, which is that Arps use of statistics is bogus, nowhere has BAC or the brief cameo star appeaarnce of Zeuzzz, addressed the issue of sample bias and normative distributions.

In fact Arp pretty much admits his selection bias.

 

[Dancing David]

What, and miss out on another page or five of posts I can add to the 'Questions BAC has run away from' thread (or whatever it's called)?

You of all JREF forum members should surely appreciate the value of holding people to account for what they write, shouldn't you*? I mean, especially given the venom, bombast, ridicule, and more heaped upon those who didn't go along with BAC's certainty over what we might call, in shorthand, 'the Arp Karlsson peak quasars'?

And surely if there's anything to 'Arp's statistics', given how dramatically it would change not only astronomy but also physics, don't you think it worthwhile to run it to ground (and show just how full of errors of many different kinds this idea/these ideas is/are ... or not)?

* at least in this section of the forum, and on matters with clear scientific content.

I feel that running it to ground was the point of the thread!

And what is even more amazing is that BAC has not presented 'a good source on Karlsson peaks' as I asked a while back. Perhaps there is none.

DUH, luminosity and redshift have an relationship.


So if you cut QSOs into slices by magnitude vs. redshift, what do you get?

Confirmation that the redshift is related to luminosity?