Celestia Forums

Hi,

yes, it was a seemingly small issue that kept me quite busy, recently: the precise alignment of the MilkyWay using my standard quaternionic rotations and the "new" IAU MilkyWay coordinates of 1959.

The origin of the problam came from a small 1.5 degree sideways discrepancy with Selden's (semi-empirical) graticule alignment of the galactic plane...Since my procedure really went back to the basics, I wanted to clear this up before committing my galaxy upgrade to CVS.

The reason for the discrepancy was tricky, as usual...Now that I solved it, let me show you how beautifully consistent everything comes out. As an independent check on the alignment, I followed and expanded Selden's method of directly generating & marking accurate points in the galactic plane and above by using this professional-level coordinate calculator:

http://cxc.harvard.edu/toolkit/precess.jsp

of the Chandra X-ray Center.

In the following top and sideways views you see the generated points in form of yellow crosses and one green cross that sits above the galactic plane just over the sun's position. You also see the MilkyWay as aligned via quaternion rotations from the IAU data ONLY. Moreover, I have added the pulsar data from Selden as a crosscheck on the MilkyWay position angle of 122.932 degrees. The matching of the pulsar locations with the MW arms has now even been considerable improved!

Have a click for kingsize (important!)!

The green cross marks the sun's position in Orion's arm of the MW. The line from there to the center of the MW corresponds to zero galactic longitude.


The second image shows sideways views of the alignment quality where the top and bottom images differ by a 90 degree rotation within the galactic plane. Look for the sun's position...

So all in all the alignment is PERFECT now!

This is also an important crosscheck on my PERL orientation routines that are used to align 10000+ galaxies!

My galaxy database upgrade was a lot of work...

I have almost completely rewritten the ~1300 line PERL extraction script, since now many more professional catalogs have been merged to determine distances and other crucial galaxy data. Notably, I have now implemented the complete LOCAL GROUP of galaxies from a new catalog on "Galaxies in Local Group Volume (d < 10 Mpc)" and from the new "NED-1D Distance Data Base". Also the complete zcat redshift catalog and the NED batch service has been exploited...

So here is the final account:


While previously, only about 60% of my galaxies in deepsky.dsc involved reliable distance measurements now I am VERY close to 100%. Trara! Trara!


Summary Output:
-----------------------
10937 Galaxies in Total
10032 with good distance data

My new PERL extraction script now can also list all galaxies of the local group for people to check. The output should look like this:


So now it's all up to you! We need to test the many new galaxies and I hope you will help...Finally myself and other people have written useful .celx scripts, that one may use for displaying the local group and other useful features.

Here are mine

http://www.celestiaproject.net/~t00fri/images/z-dist.celx
http://www.celestiaproject.net/~t00fri/images/mark-lg.celx

The first z-dist.celx script marks ALL galaxies in my new catalog and looks at them from a large distance.
The galaxy dots get increasingly read the higher their redshift! You may use the command-line to select any galaxy in the lot. It will then be marked as a green dot. Initially the MilkyWay is selected. There may be some recent changes or adaptations necessary, since Vincent has been industrious making all sorts of celx changes in CVS. But basically I checked that the scripts work fine.

The second script, mark-lg.celx marks ALL members of the local group of galaxies. By turning on the labels and increasing the sensitivity (] key) you get a nice overview of what is new as to the local group of galaxies.

Probably tonight, I shall commit the whole lot to CVS so you can download and test everything (=> DONE!)!

Enjoy,
Bye Fridger

Fridger,

even if I already told you before, I must insist that this work is probably the most important one made for Celestia to date (at least, on my personal view). There's no way I can thank you enough for this VERY complex and laborious task ! This is simply just amazing !

I raise my hat to you !

Cham wrote:Fridger,

even if I already told you before, I must insist that this work is probably the most important one made for Celestia to date (at least, on my personal view). There's no way I can thank you enough for this VERY complex and laborious task ! This is simply just amazing !

I raise my hat to you !

Martin,

of course everyone enjoys friendly words...yet now we got to eliminate together all the small buggies that may still lurk in the "dark"

Anyway thanks for your appreciation!

Cheers,
Fridger

Very impressive! I'm looking forward to testing this tomorrow.

--Chris

Just transferred everything to CVS. Let's go and test things out....

Bye Fridger

As we discussed and agreed previously, it's important that my PERL scripts will be part of the Windows distribution, such that there is a human readable documentation of the data used.

I thought it might also be useful to include some .celx scripts that serve to mark and display the galaxy data.

It would be great if Vincent could overhaul my two scripts above in the light of his recent CELX changes. Then I would simply add them to src/tools/galaxies in CVS and

+++++++++++++++++++
Chris hopefully will not forget to include them in the distribution as well.
+++++++++++++++++++

Bye Fridger

t00fri wrote:While previously, only about 60% of my galaxies in deepsky.dsc involved reliable distance measurements now I am VERY close to 100%. Trara! Trara!

t00fri wrote:As we discussed and agreed previously, it's important that my PERL scripts will be part of the Windows distribution, such that there is a human readable documentation of the data used.

I thought it might also be useful to include some .celx scripts that serve to mark and display the galaxy data.

It would be great if Vincent could overhaul my two scripts above in the light of his recent CELX changes. Then I would simply add them to src/tools/galaxies in CVS and

+++++++++++++++++++
Chris hopefully will not forget to include them in the distribution as well.
+++++++++++++++++++

It's my opinion that a user sophisticated enough to make use of the Perl scripts will also know how to get them from SourceForge However, I have no objection to adding them to the Windows package, and the celx scripts would be most useful as well. I think the best way to ensure all the relevant scripts are included would be for you to edit the Windows install script yourself and add all of the required perl and celx files.

We should think about what other celx scripts might be useful to include in the release, and possibly eliminate some of the less useful ones present in the scripts directory.

--Chris

t00fri wrote:It would be great if Vincent could overhaul my two scripts above in the light of his recent CELX changes. Then I would simply add them to src/tools/galaxies in CVS and

+++++++++++++++++++
Chris hopefully will not forget to include them in the distribution as well.
+++++++++++++++++++

Bye Fridger

Of course, Fridger, with great pleasure. I'll send them to you via email once I'm finished.
And congratulation for this fantastic work !

chris wrote:It's my opinion that a user sophisticated enough to make use of the Perl scripts will also know how to get them from SourceForge
...
--Chris

Chris,

while your statement is certainly right, there is nevertheless a CATCH in the argument!

+++++++++++++++++++++++
The catch is that those sophisticated Windows users have NO IDEA that these scripts are available in the source code!
+++++++++++++++++++++++


Windows users being provided with a functioning /binary/ distribution, hardly go on investigating what else might be available somewhere.

This is really a crucial point I think, that you have overlooked since a long time...

I also think we should contemplate a collection of other useful scripts to be included in the distribution!

Bye Fridger

Vincent wrote:

t00fri wrote:It would be great if Vincent could overhaul my two scripts above in the light of his recent CELX changes. Then I would simply add them to src/tools/galaxies in CVS and

+++++++++++++++++++
Chris hopefully will not forget to include them in the distribution as well.
+++++++++++++++++++

Bye Fridger

Of course, Fridger, with great pleasure. I'll send them to you via email once I'm finished.
And congratulation for this fantastic work !

Thanks, Vincent,

that's great...
Bye Fridger

I certainly think that we should include Selden's Pulsar mark script! It's just great...

Bye Fridger

t00fri wrote:

chris wrote:It's my opinion that a user sophisticated enough to make use of the Perl scripts will also know how to get them from SourceForge
...
--Chris

Chris,

while your statement is certainly right, there is nevertheless a CATCH in the argument!

+++++++++++++++++++++++
The catch is that those sophisticated Windows users have NO IDEA that these scripts are available in the source code!
+++++++++++++++++++++++

Let's put them in the distribution. They certainly won't increase the size of the package by any significant amount. Would you take on the task of adding them to the Windows install script?

I certainly think that we should include Selden's Pulsar mark script! It's just great...

Agreed--it should be included.

--Chris

The Pulsar Addon was created in December '04. I'm sure they've added a few to the database since then.

I won't be able to have an update ready for this weekend, though: I have other commitments.

[edit]
The existing Addon has 1509 pulsars.
The ATNF database currently has 1771, almost 12% more.
[/edit]

Kudos for your remarkable work Fridger. Very impressive.
Cosmologic Celestia starts now officially.
Thank you very much.

selden wrote:The Pulsar Addon was created in December '04. I'm sure they've added a few to the database since then.

I won't be able to have an update ready for this weekend, though: I have other commitments.

[edit]
The existing Addon has 1509 pulsars.
The ATNF database currently has 1771, almost 12% more.
[/edit]

I have never looked into that catalog, but I would guess that extracting the pulsar data is probably a most simple task for PERL freaks like myself? Your mark script for the individual pulsars can also be generated with a few lines of PERL code. Selden? What data did you extract?

RAJD, DecJD, NameJ Period (P0?) and what density parameter did you take?

But I have a job already for the weekend: Inno setup...

Let's see what can be done.

Bye Fridger

Be cautious with that database, there are several duplicates of the same pulsar with several names. At the time Selden extracted the database, I tested the results extensively and have found several "minor" problems (duplicates). Some pulsars even have a very wrong distance, or can be found at several places (with some variations in the names).

Well there are plenty of other things for me to do before the 1.5.0pre3 deadline tomorrow. So I will at best be able to upgrade the pulsars for 1.5.0final, despite this being a really simple PERL task (just one catalog, not 10 to merge...).

Also I surely would make this and that different. Since it is Selden's add-on I do not want to interfere, however, ....

Anyway, I think this add-on script should be part of Celestia, since it is both most instructive and serves as the best cross check on the arm structure of our MW template....Moreover it is small...

Bye Fridger

Unfortunately, I can't find the programs I wrote to translate the ATNF database file into an STC catalog and scripts, so a lot more work would be involved than I was expecting. Due to other commitments, I won't be able to work on this project any time soon. Sorry.

If someone else has the time to generate appropriate data files, please do so.

Fridger,

your CELX scripts don't work anymore on my system with the latest version of Celestia from Dirkpitt. They were working perfectly before, with my own builds from CVS. All it does now is unmark everything, then nothing. A vast part of the cosmos is shown without any label. Something is broken in Dirkpitt's version.

There's something weird with M 54 and Sagittarius dSph. They are both PERFECTLY aligned with Sol and I wonder if they are both one single object shown with different distances.

I'm also wondering if IC 3658 should be part of the local group. It's so close of the other members.

Here's my CEL script to show the local group (a simple variant of yours) :