Skip to main content

TR Memescape

  • Talk Rational: The aspies on this board post long posts about being aspie, and the ADHDers can't read to the end of them.

Topic: Bootstrapping (Read 1128 times) previous topic - next topic

0 Members and 1 Guest are viewing this topic.
  • socrates1
Re: Bootstrapping
Reply #25
http://www.ivpp.cas.cn/qt/papers/201403/P020140314389417822583.pdf (2011)
An Archaeopteryx-like theropod from China and the origin of Avialae
Xing Xu1,2, Hailu You3 , Kai Du4 & Fenglu Han2

https://4.bp.blogspot.com/-2GFhO-iuwIA/WIDkpfaHe7I/AAAAAAAAAQs/flD8HdBGpFQFoELqdEjEBE3HEul-5MErgCLcB/s1600/polytomy.png
Figure S9. Bootstrap values for the clades recovered by our analysis of coelurosaurian phylogeny (only clades with bootstrap values greater than 50% are shown).

A polytomy like that in Figure S9 means that we cannot tell the relationship between the dozens of branches.

The bottom line is that cladistic analysis itself does not allow us to conclude anything about the relationship between the dozens of branches.
If we want to come to any conclusions about those relationships, cladistic analysis itself is no help.

Here is confirmation of this point from an expert in the field:

"Regarding your last message, yes, you are right: if there is a polytomy we cannot tell with good support what the internal relationships between the branches inside that group are."

I am not intending to argue this point. If you do not understand it, you will need to study the subject yourself.



So if someone says that cladistic analysis shows relationships between the dozens of branches they are incorrect.

This is significant, because most cladograms that have been published, incorrectly show relationships between the branches of that huge polytomy. That is because they ignore the facts about the bootstrap/jackknife values.

  • Faid
Re: Bootstrapping
Reply #26
Still here?

You must be very lonely.
Who even made the rule that we cannot group ducks and fish together for the simple reason that they are both aquatic? If I want to group them that way and it serves my purpose then I can jolly well do it however I want to and it is still a nested hierarchy and you can't tell me that it's not.

Re: Bootstrapping
Reply #27
http://www.ivpp.cas.cn/qt/papers/201403/P020140314389417822583.pdf (2011)
An Archaeopteryx-like theropod from China and the origin of Avialae
Xing Xu1,2, Hailu You3 , Kai Du4 & Fenglu Han2

https://4.bp.blogspot.com/-2GFhO-iuwIA/WIDkpfaHe7I/AAAAAAAAAQs/flD8HdBGpFQFoELqdEjEBE3HEul-5MErgCLcB/s1600/polytomy.png
Figure S9. Bootstrap values for the clades recovered by our analysis of coelurosaurian phylogeny (only clades with bootstrap values greater than 50% are shown).

And here's the quote from whatever expert Socrates wasted the time of.

"Regarding your last message, yes, you are right: if there is a polytomy we cannot tell with good support what the internal relationships between the branches inside that group are."

There's certainly no polytomy in grouping everything on the branch containing Haplocherius at the top to Anserimimus at the bottom (including several uncontested theropod dinosaurs such as Tyrannosaurus BTW) as more closely related to each other than to the uncontested theropod Sinraptor (with a slamdunk bootstrap value of 97).  There's also no polytomy in grouping that group of theropods together with Sinraptor as being more closely related to each other than they are to the uncontested theropod Allosaurus (with an ever greater slamdunk bootstrap value of 100).

It's a lock, folks.  Everything in that cladogram from Haplocherius to Anserimimus is a theropod dinosaur, regardless of their internal grouping.

Birds are dinosaurs.  End of the line.

But Socrates, feel free to go on and on about how the polytomy means we can't be too confident about the internal relationships between the branches of theropod dinosaurs inside the clade containing Haplocherius to Anserimimus.
  • Last Edit: June 17, 2017, 11:41:50 AM by Dean W

  • socrates1
Re: Bootstrapping
Reply #28
http://www.ivpp.cas.cn/qt/papers/201403/P020140314389417822583.pdf (2011)
An Archaeopteryx-like theropod from China and the origin of Avialae
Xing Xu1,2, Hailu You3 , Kai Du4 & Fenglu Han2

https://4.bp.blogspot.com/-2GFhO-iuwIA/WIDkpfaHe7I/AAAAAAAAAQs/flD8HdBGpFQFoELqdEjEBE3HEul-5MErgCLcB/s1600/polytomy.png
Figure S9. Bootstrap values for the clades recovered by our analysis of coelurosaurian phylogeny (only clades with bootstrap values greater than 50% are shown).

A polytomy like that in Figure S9 means that we cannot tell the relationship between the dozens of branches.

The bottom line is that cladistic analysis itself does not allow us to conclude anything about the relationship between the dozens of branches.
If we want to come to any conclusions about those relationships, cladistic analysis itself is no help.

Here is confirmation of this point from an expert in the field:

"Regarding your last message, yes, you are right: if there is a polytomy we cannot tell with good support what the internal relationships between the branches inside that group are."

I am not intending to argue this point. If you do not understand it, you will need to study the subject yourself.



So if someone says that cladistic analysis shows relationships between the dozens of branches they are incorrect.

This is significant, because most cladograms that have been published, incorrectly show relationships between the branches of that huge polytomy. That is because they ignore the facts about the bootstrap/jackknife values.


The Brusatte et al (2014) cladogram was done later and is better than the Xu et al (2011) study because it shows the tyrannosauroids separate from the huge polytomy.
The node of the huge polytomy does not have a name, so I will label it "Euparaves" which is a good name for it since it is actually basal Paraves.
Whatever you call that node, it shows that there is no evidence that links the tyrannosauroids and that node ("Euparaves").
  • Last Edit: June 17, 2017, 12:30:10 PM by socrates1

Re: Bootstrapping
Reply #29
lol this thred

Re: Bootstrapping
Reply #30
If someone with actual real-world experience of doing cladistics who understands the statistics involved would like to contribute, I for one, would be grateful to have any errors I may have made pointed out. I get that you'd just be writing a Cladistics 101 course, and aint nobody got time for that, but its been a long time since I did it, and the statistical methodology of checking your results wasn't really a part of it. 
Why do I bother?

  • Doobie Keebler
  • Ridiculous Callipygous
Re: Bootstrapping
Reply #31
I have perhaps gone too fast. It has probably not sunk in that in this case cladistic analysis cannot help in determining the relationship of the branches in the polytomy.

You have not gone too fast. Your deep seated narcissism has you stuck in a fallacious loop where just like every other crank on the internet you mistakenly believe that if you attack something else, rather bother to support your own idea with evidence, your idea wins.

"Everyone here" recognizes this. Except you.

That is a blockbuster.

No, this is a blockbuster. And it is as empty and devoid of substance as the support for your wankery.

But please do continue to tell us more about how some other idea is wrong, sock-o.  ::)

  • Last Edit: June 17, 2017, 02:18:19 PM by Doobie Keebler
"I'm over 70 and have never seen such , arrogance, incompetence and Ill -intentions as this President and his aids."    The Dotard     (posted 12 days after his 68th birthday)

  • Faid
Re: Bootstrapping
Reply #32
The "better" Brusatte et al (2014) cladogram has Tyrannosauroidea AND all maniraptoriformes WITHIN the coelurosauria clade, with 94% support.

We accept your apology.

Anything else?

If not, bye.
Who even made the rule that we cannot group ducks and fish together for the simple reason that they are both aquatic? If I want to group them that way and it serves my purpose then I can jolly well do it however I want to and it is still a nested hierarchy and you can't tell me that it's not.

  • socrates1
Re: Bootstrapping
Reply #33
http://www.ivpp.cas.cn/qt/papers/201403/P020140314389417822583.pdf (2011)
An Archaeopteryx-like theropod from China and the origin of Avialae
Xing Xu1,2, Hailu You3 , Kai Du4 & Fenglu Han2

https://4.bp.blogspot.com/-2GFhO-iuwIA/WIDkpfaHe7I/AAAAAAAAAQs/flD8HdBGpFQFoELqdEjEBE3HEul-5MErgCLcB/s1600/polytomy.png
Figure S9. Bootstrap values for the clades recovered by our analysis of coelurosaurian phylogeny (only clades with bootstrap values greater than 50% are shown).

A polytomy like that in Figure S9 means that we cannot tell the relationship between the dozens of branches.

The bottom line is that cladistic analysis itself does not allow us to conclude anything about the relationship between the dozens of branches.
If we want to come to any conclusions about those relationships, cladistic analysis itself is no help.

Here is confirmation of this point from an expert in the field:

"Regarding your last message, yes, you are right: if there is a polytomy we cannot tell with good support what the internal relationships between the branches inside that group are."

I am not intending to argue this point. If you do not understand it, you will need to study the subject yourself.



So if someone says that cladistic analysis shows relationships between the dozens of branches they are incorrect.

This is significant, because most cladograms that have been published, incorrectly show relationships between the branches of that huge polytomy. That is because they ignore the facts about the bootstrap/jackknife values.


The Brusatte et al (2014) cladogram was done later and is better than the Xu et al (2011) study because it shows the tyrannosauroids separate from the huge polytomy.
The node of the huge polytomy does not have a name, so I will label it "Euparaves" which is a good name for it since it is actually basal Paraves.
Whatever you call that node, it shows that there is no evidence that links the tyrannosauroids and that node ("Euparaves").

The cladistic analyses show that a dinosaur to bird hypothesis requires a huge saltation between creatures like tyrannosauroids and Euparaves.
  • Last Edit: June 17, 2017, 03:25:35 PM by socrates1

  • VoxRat
  • wtactualf
Re: Bootstrapping
Reply #34
Wow!
7 deep!
It's a new record.
:golfclap:

Let's see how deep "Socrates" will nest his self-quotes! :popcorn:
"I understand Donald Trump better than many people because I really am a lot like him." - Dave Hawkins

Re: Bootstrapping
Reply #35

The Brusatte et al (2014) cladogram was done later and is better than the Xu et al (2011) study...


Socrates, several of us (but curiously not you) have noted that Brusatte et al. found Coelurosauria (including paravians) to be a clade with a jackknife value of 94, while Xu et al. found Coelurosauria (including paravians) to be a clade with a bootstrap value of 97.  Perhaps you prefer Brusatte et al. because 94 is less than 97.  But to any objective observer, the evidence indicates that Coelurosauria (including paravians) is a strongly supported clade by either study.  That objective observer would no doubt conclude that the evidence of both studies supports paravians being deeply nested within theropod dinosaurs and only distantly related to pterosaurs.

I think our guests would love for you to discuss this.  In fact, should you continue to shy away from discussing this elephant in the room and focus only on uncertainty regarding subclades within Coelurosauria, they might conclude that you are simply an evasive crackpot who is incensed that his pet theory has been demolished by either study.  We wouldn't want that, would we?

  • socrates1
Re: Bootstrapping
Reply #36
Quote
The Brusatte et al (2014) cladogram was done later and is better than the Xu et al (2011) study because it shows the tyrannosauroids separate from the huge polytomy.
The node of the huge polytomy does not have a name, so I will label it "Euparaves" which is a good name for it since it is actually basal Paraves.
Whatever you call that node, it shows that there is no evidence that links the tyrannosauroids and that node ("Euparaves").
The cladistic analyses show that a dinosaur to bird hypothesis requires a huge saltation between creatures like tyrannosauroids and Euparaves.

The interesting work begins AFTER the acknowledgement of this fact.
In the published literature nobody has tried to explain the huge saltation, because they have not even acknowledged the huge polytomy, even though they document it in the supplementary information sections of their studies.
Of course, the huge saltation can never be explained.
  • Last Edit: June 17, 2017, 05:44:48 PM by socrates1

Re: Bootstrapping
Reply #37
Guests, for reference, Socrates is referring to a polytomy WITHIN the strongly supported clade of coelurosaurian theropod dinosaurs (including paravians) as found by both the Brusatte et al. and Xu et al. studies.  That's the polytomy under discussion.

Socrates, carry on if you please.

ETA: And guests, neither paper mentions saltation of any kind.  Socrates is making that up.
  • Last Edit: June 17, 2017, 06:15:28 PM by Dean W

  • Doobie Keebler
  • Ridiculous Callipygous
Re: Bootstrapping
Reply #38
Is this when sock-o's accusations of "everyone here" pretending to not understand begin?

I forget.
"I'm over 70 and have never seen such , arrogance, incompetence and Ill -intentions as this President and his aids."    The Dotard     (posted 12 days after his 68th birthday)

Re: Bootstrapping
Reply #39
The strong support of Coelurosauria (including paravians) as a clade of theropod dinosaur in every study brought to light is something our guests will never see Socrates address.  His silence on this matter is tacit admission that he cannot dispute it.  Instead, he childishly pretends unresolved relationships within coelurosaurian theropod dinosaurs (including paravians) will persuade our guests that paravians should be removed from Coelurosauria and grouped with pterosaurs, a truly ludicrous proposal unsupported by any evidence presented here or on his moribund blog.

  • socrates1
Re: Bootstrapping
Reply #40
Quote
The Brusatte et al (2014) cladogram was done later and is better than the Xu et al (2011) study because it shows the tyrannosauroids separate from the huge polytomy.
The node of the huge polytomy does not have a name, so I will label it "Euparaves" which is a good name for it since it is actually basal Paraves.
Whatever you call that node, it shows that there is no evidence that links the tyrannosauroids and that node ("Euparaves").
The cladistic analyses show that a dinosaur to bird hypothesis requires a huge saltation between creatures like tyrannosauroids and Euparaves.

The interesting work begins AFTER the acknowledgement of this fact.
In the published literature nobody has tried to explain the huge saltation, because they have not even acknowledged the huge polytomy, even though they document it in the supplementary information sections of their studies.
Of course, the huge saltation can never be explained.

If they acknowledged the huge saltation, that would be the collapse of the dinosaur to bird hypothesis. Euparaves is not a node in a dinosaur to bird phylogeny. The 65 jackknife value indicates that tyrannosauroids are not related to Euparaves.

  • Monad
Re: Bootstrapping
Reply #41
lol no it doesn't. You don't understand any of this so stop pretending you do.

  • socrates1
Re: Bootstrapping
Reply #42
Quote
The Brusatte et al (2014) cladogram was done later and is better than the Xu et al (2011) study because it shows the tyrannosauroids separate from the huge polytomy.
The node of the huge polytomy does not have a name, so I will label it "Euparaves" which is a good name for it since it is actually basal Paraves.
Whatever you call that node, it shows that there is no evidence that links the tyrannosauroids and that node ("Euparaves").
The cladistic analyses show that a dinosaur to bird hypothesis requires a huge saltation between creatures like tyrannosauroids and Euparaves.

The interesting work begins AFTER the acknowledgement of this fact.
In the published literature nobody has tried to explain the huge saltation, because they have not even acknowledged the huge polytomy, even though they document it in the supplementary information sections of their studies.
Of course, the huge saltation can never be explained.

If they acknowledged the huge saltation, that would be the collapse of the dinosaur to bird hypothesis. Euparaves is not a node in a dinosaur to bird phylogeny. The 65 jackknife value indicates that tyrannosauroids are not related to Euparaves.

The alternatives are "thecodonts" or pterosaurs. And the thecodont hypothesis suffers from the same problems as the dinosaur to bird hypothesis. Keep in mind that the 65 jackknife value indicates that tyrannosauroids are not related to Euparaves.
  • Last Edit: June 18, 2017, 04:18:38 AM by socrates1

  • Faid
Re: Bootstrapping
Reply #43

If I acknowledged the huge saltation within the coelurosauria clade, that would be the confirmation of the dinosaur to bird hypothesis, since "Euparaves" are included with 94% support in coelurosaur dinosaurs. Euparaves is not a node in a dinosaur to bird phylogeny, since it is included in the 94% jackknife- supported coleurosauria clade. The 65 jackknife value indicates that tyrannosauroids are not related to Euparaves. The form of relation is not resolved, but the fact that they are both included in the 94-jackknife-value coleurosauria clade confims beyond doubt that the relation exists, as any non-moronic person can comprehend.
Fixed your post.

Thank me later.

Anything else?
Who even made the rule that we cannot group ducks and fish together for the simple reason that they are both aquatic? If I want to group them that way and it serves my purpose then I can jolly well do it however I want to and it is still a nested hierarchy and you can't tell me that it's not.

  • socrates1
Re: Bootstrapping
Reply #44
Quote
The Brusatte et al (2014) cladogram was done later and is better than the Xu et al (2011) study because it shows the tyrannosauroids separate from the huge polytomy.
The node of the huge polytomy does not have a name, so I will label it "Euparaves" which is a good name for it since it is actually basal Paraves.
Whatever you call that node, it shows that there is no evidence that links the tyrannosauroids and that node ("Euparaves").
The cladistic analyses show that a dinosaur to bird hypothesis requires a huge saltation between creatures like tyrannosauroids and Euparaves.

The interesting work begins AFTER the acknowledgement of this fact.
In the published literature nobody has tried to explain the huge saltation, because they have not even acknowledged the huge polytomy, even though they document it in the supplementary information sections of their studies.
Of course, the huge saltation can never be explained.

If they acknowledged the huge saltation, that would be the collapse of the dinosaur to bird hypothesis. Euparaves is not a node in a dinosaur to bird phylogeny. The 65 jackknife value indicates that tyrannosauroids are not related to Euparaves.

The alternatives are "thecodonts" or pterosaurs. And the thecodont hypothesis suffers from the same problems as the dinosaur to bird hypothesis. Keep in mind that the 65 jackknife value indicates that tyrannosauroids are not related to Euparaves.

And it attains a value of 65 probably only because dinosaurs share archosaur symplesiomorphic characters with Euparaves. Support value calculations (bootstrap/jackknife) use characters and not synapomorphies.
  • Last Edit: June 18, 2017, 04:38:49 AM by socrates1

  • Faid
Re: Bootstrapping
Reply #45
Quote
The Brusatte et al (2014) cladogram was done later and is better than the Xu et al (2011) study because it shows the tyrannosauroids separate from the huge polytomy.
The node of the huge polytomy does not have a name, so I will label it "Euparaves" which is a good name for it since it is actually basal Paraves.
Whatever you call that node, it shows that there is no evidence that links the tyrannosauroids and that node ("Euparaves").
The cladistic analyses show that a dinosaur to bird hypothesis requires a huge saltation between creatures like tyrannosauroids and Euparaves.

The interesting work begins AFTER the acknowledgement of this fact.
In the published literature nobody has tried to explain the huge saltation, because they have not even acknowledged the huge polytomy, even though they document it in the supplementary information sections of their studies.
Of course, the huge saltation can never be explained.

If they acknowledged the huge saltation, that would be the collapse of the dinosaur to bird hypothesis. Euparaves is not a node in a dinosaur to bird phylogeny. The 65 jackknife value indicates that tyrannosauroids are not related to Euparaves.

The alternatives are "thecodonts" or pterosaurs. And the thecodont hypothesis suffers from the same problems as the dinosaur to bird hypothesis. Keep in mind that the 65 jackknife value indicates that tyrannosauroids are not related to Euparaves.
Keep in mind that the 94 jackknife value indicates that both tyrannosaurids and "Euparaves" are coelurosaur dinosaurs. Therefore, for your pterosaur hypothesis to even begin to stand up, one would have to show that pterosaurs are ALSO more basal forms of coelurosaur dinosaurs. So one would have to produce a strongly supported (>70, as per your accepted criteria) node that includes pterosaurs within coleurosaur dinos,  as well as a more derived, equally strong-supported node that includes "Euparaves" and pterosaurs.

I'm sure that won't be a problem for a genius such as you.

Good luck.
Who even made the rule that we cannot group ducks and fish together for the simple reason that they are both aquatic? If I want to group them that way and it serves my purpose then I can jolly well do it however I want to and it is still a nested hierarchy and you can't tell me that it's not.

  • Faid
Re: Bootstrapping
Reply #46
And it attains a value of 65 probably only because dinosaurs share archosaur symplesiomorphic characters with Euparaves. And support value calculations use characters and not synapomorphies.
So why does the coelurosauria node "attain" a value of 94 with "Euparaves" included?

Not that I expect you to grow the balls to address that.

End of the line.
Who even made the rule that we cannot group ducks and fish together for the simple reason that they are both aquatic? If I want to group them that way and it serves my purpose then I can jolly well do it however I want to and it is still a nested hierarchy and you can't tell me that it's not.

  • socrates1
Re: Bootstrapping
Reply #47
Quote
The Brusatte et al (2014) cladogram was done later and is better than the Xu et al (2011) study because it shows the tyrannosauroids separate from the huge polytomy.
The node of the huge polytomy does not have a name, so I will label it "Euparaves" which is a good name for it since it is actually basal Paraves.
Whatever you call that node, it shows that there is no evidence that links the tyrannosauroids and that node ("Euparaves").
The cladistic analyses show that a dinosaur to bird hypothesis requires a huge saltation between creatures like tyrannosauroids and Euparaves.

The interesting work begins AFTER the acknowledgement of this fact.
In the published literature nobody has tried to explain the huge saltation, because they have not even acknowledged the huge polytomy, even though they document it in the supplementary information sections of their studies.
Of course, the huge saltation can never be explained.

If they acknowledged the huge saltation, that would be the collapse of the dinosaur to bird hypothesis. Euparaves is not a node in a dinosaur to bird phylogeny. The 65 jackknife value indicates that tyrannosauroids are not related to Euparaves.

The alternatives are "thecodonts" or pterosaurs. And the thecodont hypothesis suffers from the same problems as the dinosaur to bird hypothesis. Keep in mind that the 65 jackknife value indicates that tyrannosauroids are not related to Euparaves.

And it attains a value of 65 probably only because dinosaurs share archosaur symplesiomorphic characters with Euparaves. Support value calculations (bootstrap/jackknife) use characters and not synapomorphies.

The same logic applies to the high support values the earlier nodes show.

  • Faid
Re: Bootstrapping
Reply #48
Quote
"Euparaves" are included with 94% support in coelurosaur dinosaurs.
Worth repeating.

Poor "socrates".
Who even made the rule that we cannot group ducks and fish together for the simple reason that they are both aquatic? If I want to group them that way and it serves my purpose then I can jolly well do it however I want to and it is still a nested hierarchy and you can't tell me that it's not.

  • Faid
Re: Bootstrapping
Reply #49
The same logic applies to the high support values the earlier nodes show.
:rofl:

"Low bootstrap/jackknife values indicate that there is no relation, and they are just archosaur symplesiomorphies. Oh and the same goes for HIGH bootstrap/jackknife values".

Poor "socrates". He's all over the shop.
Who even made the rule that we cannot group ducks and fish together for the simple reason that they are both aquatic? If I want to group them that way and it serves my purpose then I can jolly well do it however I want to and it is still a nested hierarchy and you can't tell me that it's not.