Dear Michael Borland
Hello. I'm master course student in S.Korea at KNU.
Nowadays I study about frequency map analysis. But Still I don't know about command for FMA, such as &frequency_map.
So, If you have some example about FMA, Please could you attach the file to me?
And I have one more question. How to plot tune shift with resonance lines?
I hear about FMA has bugs in 32 bit computer. Is it fixed?
Sincerely yours
From Hwang Ji-Gwang
frequency map analysis
Moderators: cyao, michael_borland
-
michael_borland
- Posts: 1959
- Joined: 19 May 2008, 09:33
- Location: Argonne National Laboratory
- Contact:
Re: frequency map analysis
Ji-Gwang,
There is an FMA example in the examples file which you can find here
http://www.aps.anl.gov/Accelerator_Syst ... ml#elegant
Once you download and unpack the file, look in PAR/frequencyMap
To plot the resonance lines, use the program sddsresdiag to generate a resonance diagram in a file.
The FMA bug on 32 bit was fixed several releases ago. If you have the latest release, it is fixed.
--Michael
There is an FMA example in the examples file which you can find here
http://www.aps.anl.gov/Accelerator_Syst ... ml#elegant
Once you download and unpack the file, look in PAR/frequencyMap
To plot the resonance lines, use the program sddsresdiag to generate a resonance diagram in a file.
The FMA bug on 32 bit was fixed several releases ago. If you have the latest release, it is fixed.
--Michael
-
Hwang, Ji-Gwang
- Posts: 42
- Joined: 21 Jan 2010, 02:27
Re: frequency map analysis
Dear Michael Borland
Hello. I try to run the example in PAR/frequencyMap directory through cygwin using windows mechine.
And I get a some result. But I have more questions.
I don't know why axis of the plot of nux and nuy is fractional values?
I think it means that It represent to fraction value of tune of bare lattce. It is right?
And I want to make more small squares or dot in polt.
Actually I change the command "sacle=3->scale=1" in the track file.
So, I get more small squares. but, It is not sufficient to me.
Could you tell me more hints?
From Hwang, Ji-Gwang
Hello. I try to run the example in PAR/frequencyMap directory through cygwin using windows mechine.
And I get a some result. But I have more questions.
I don't know why axis of the plot of nux and nuy is fractional values?
I think it means that It represent to fraction value of tune of bare lattce. It is right?
And I want to make more small squares or dot in polt.
Actually I change the command "sacle=3->scale=1" in the track file.
So, I get more small squares. but, It is not sufficient to me.
Could you tell me more hints?
From Hwang, Ji-Gwang
- Attachments
-
- mpl.png (6.63 KiB) Viewed 14469 times
-
michael_borland
- Posts: 1959
- Joined: 19 May 2008, 09:33
- Location: Argonne National Laboratory
- Contact:
Re: frequency map analysis
Ji-Gwang,
FMA, like FFT, can only determine the fractional part of the tune. Try this:
sddsresdiag resdiag.sdds
sddsplot -col=nux,nuy parTrack.fma -graph=dot,vary -order=spect -split=col=x -col=nux,nuy resdiag.sdds -sever
This will plot the tunes color-coded by initial x amplitude along with a resonance diagram.
--Michael
FMA, like FFT, can only determine the fractional part of the tune. Try this:
sddsresdiag resdiag.sdds
sddsplot -col=nux,nuy parTrack.fma -graph=dot,vary -order=spect -split=col=x -col=nux,nuy resdiag.sdds -sever
This will plot the tunes color-coded by initial x amplitude along with a resonance diagram.
--Michael
-
Hwang, Ji-Gwang
- Posts: 42
- Joined: 21 Jan 2010, 02:27
Re: frequency map analysis
Dear Michael Borland
Thank you for your kind reply. Your reply is very helpful for me.
I get the some result. So, I attached it.
But, I still have one question.
In the figure of result, One point aroud nux=0.9, nuy=0.3 is very far from order points.
I don't know that what means of this point?
I attached other result of FMA get from PLS. This figure look like strange at near the nux=0 and nux=1.
It is right result?
From Hwang, Ji-Gwang
Thank you for your kind reply. Your reply is very helpful for me.
I get the some result. So, I attached it.
But, I still have one question.
In the figure of result, One point aroud nux=0.9, nuy=0.3 is very far from order points.
I don't know that what means of this point?
I attached other result of FMA get from PLS. This figure look like strange at near the nux=0 and nux=1.
It is right result?
From Hwang, Ji-Gwang
- Attachments
-
-
- mpl.png (11.72 KiB) Viewed 14456 times
-
michael_borland
- Posts: 1959
- Joined: 19 May 2008, 09:33
- Location: Argonne National Laboratory
- Contact:
Re: frequency map analysis
Ji-Gwang,
The particle with strange tune values is probably near a coupling resonance. If the x and y motion is strongly coupled, it can be difficult for the algorithm to find both tunes. It looks like the x tune got misidentified in that case.
Your results for PLS show a lot of tune shift with amplitude. The clumping around the 0 and 1 tune lines is probably from particles that are nearly unstable. By the way, I think you need the -order=spect option on your plots to get proper color-coding.
To say more I'd have to see your lattice file.
--Michael
The particle with strange tune values is probably near a coupling resonance. If the x and y motion is strongly coupled, it can be difficult for the algorithm to find both tunes. It looks like the x tune got misidentified in that case.
Your results for PLS show a lot of tune shift with amplitude. The clumping around the 0 and 1 tune lines is probably from particles that are nearly unstable. By the way, I think you need the -order=spect option on your plots to get proper color-coding.
To say more I'd have to see your lattice file.
--Michael
-
Hwang, Ji-Gwang
- Posts: 42
- Joined: 21 Jan 2010, 02:27
Re: frequency map analysis
Dear Michael Borland
Hello, Sorry to late reply. I attached the lattice file for FMA.
I simualte FMA under 1000 particle and 2000 turns.
Sincerely yours.
From Hwang, Ji-Gwang
Hello, Sorry to late reply. I attached the lattice file for FMA.
I simualte FMA under 1000 particle and 2000 turns.
Sincerely yours.
From Hwang, Ji-Gwang
- Attachments
-
- pls22.lte
- (2.1 KiB) Downloaded 837 times
Re: frequency map analysis
Hi everyone,
1- Is there a solution for resdiag to plot n-superperiodicity?
Example: My work has a 10 superperiodicity and tunes are nux=27.25, nuy=12.15.
After run, I got a plot like an attachment file. I read an information that is "The coordinate values are between 0 and 1. The defaults are [0.15, 0.9]x[0.15, 0.9]. in sddsplot manual for controlling the plotting region."
2- What is the main difference between &frequency_map and &tune_footprint?
Best regards,
Kahraman.
1- Is there a solution for resdiag to plot n-superperiodicity?
Example: My work has a 10 superperiodicity and tunes are nux=27.25, nuy=12.15.
After run, I got a plot like an attachment file. I read an information that is "The coordinate values are between 0 and 1. The defaults are [0.15, 0.9]x[0.15, 0.9]. in sddsplot manual for controlling the plotting region."
2- What is the main difference between &frequency_map and &tune_footprint?
Best regards,
Kahraman.
- Attachments
-
-
michael_borland
- Posts: 1959
- Joined: 19 May 2008, 09:33
- Location: Argonne National Laboratory
- Contact:
Re: frequency map analysis
Kahraman,
You can use the -superperiodicity option on sddsresdiag to make the appropriate resonance diagram.
The main reasons to use tune_footprint instead of frequency_map are
1. You want to perform optimization of the tune_footprint. You can do this using the quantities described in the manual.
2. You want a simple tune-vs-momentum scan.
--Michael
You can use the -superperiodicity option on sddsresdiag to make the appropriate resonance diagram.
The main reasons to use tune_footprint instead of frequency_map are
1. You want to perform optimization of the tune_footprint. You can do this using the quantities described in the manual.
2. You want a simple tune-vs-momentum scan.
--Michael
Re: frequency map analysis
Dear Michael,
First of all, thanks a lot for your reply.
I recognized the relation between &frequency_map and &tune_footprint. I figured out it, well. But I have a question to get more clear related with the calculation of the frequency_map. What is the background of it? I guess it runs for n-superperiodicity (full ring), doesn't it?
I am not sure because the plotting of the results of the .fma and .res files is, as if it is one superperiodicity in tune diagram. Even though we simulate it for n-superperiodicity, the resonances which are in one superperiodicity condition are still shown in graphics. See an example a few 4 th order resonances in attach file. Actually, we can see that these resonances are not effective for this example.
Best regards,
Kahraman.
First of all, thanks a lot for your reply.
I recognized the relation between &frequency_map and &tune_footprint. I figured out it, well. But I have a question to get more clear related with the calculation of the frequency_map. What is the background of it? I guess it runs for n-superperiodicity (full ring), doesn't it?
I am not sure because the plotting of the results of the .fma and .res files is, as if it is one superperiodicity in tune diagram. Even though we simulate it for n-superperiodicity, the resonances which are in one superperiodicity condition are still shown in graphics. See an example a few 4 th order resonances in attach file. Actually, we can see that these resonances are not effective for this example.
Best regards,
Kahraman.
- Attachments
-