More HBeta
Sigh, think I've just run a thing twice. Nevertheless, I shall try and run all nine simultaneously on Keter, just for fun. Lets see...
Right given up on that one. Segmentation faults all round...
Nope, didn't run a thing twice, just forgot to move a file... nevertheless, the autoruns aren't really matching with the interactive runs! Really annoying.
Oh, I see, I've been running them with a different pitch angle... That's just great...
Next time, they need be run with zero pitch angle to match the interactive stuff... most annoying...
Answers a question from yesterday, I guess. The expected answer, just not delivered in the way I desired...
So, once these are done, I rerun 8, 1 and 2 with zero pitch angle for validation, then redo the lot with a higher resolution. Currently running nine and have prepped one, leaving two to be done before doing the validation and the rest of it.
Right, 9 done, 1 going, 2 ready.
...
All nine done (yes, I spared you the details). Now have a full animation, in effect, of a deforming HBeta doppler profile, assuming a pitch angle of forty five degrees. Now set off the first of three validation runs for zero pitch angle. After that, will continue on and do the other six profiles at zero pitch angle. Once this is all done, I can grab a cosine distribution and use that. Have also used the original signal profiles and my own set of numbers to deduce correction factors to be introduced to try and get the relative intensities of the profiles correct. If I can then get one of the profiles and use it to calibrate its modelled counterpart, I'll see if I can actually model intensities through this method - assuming its any good at all... it may still not be.
...all to be done laterz...
...it is later. Will prepare the next eight runs for autorunning as background processes whilst I go home and sleep. Even with eight going, they should be done before I return tomorrow afternoon - unless someone logs on later in the hope of an empty Rimmer... it happens. Even so, that's their pigeon. After this has been done, I can start on the high res stuff. Exciting, really exciting...
Dammit - changing the pitch angle hasn't changed the autorun to the correct shape at all! Wtf is going on?
Aaaah - thermalisation level, I seeeeeee. There's a massive unshifted spike due to low level collisions. All to do with the efficiencies... sigh. Cross-sections for interactions, scattering ratios, emission efficiencies... there are so many atomic data to include in this damned thing. Still, can at least ramp up the thermalisation and give one set of results that can be used... also explains why the autoruns take so damned long...
ok, that's nine runs all happilly chugging along. One or two may even finish and get evaluated before I go... of course, I can tune the thermalisation all I want...
...time to go!
Right given up on that one. Segmentation faults all round...
Nope, didn't run a thing twice, just forgot to move a file... nevertheless, the autoruns aren't really matching with the interactive runs! Really annoying.
Oh, I see, I've been running them with a different pitch angle... That's just great...
Next time, they need be run with zero pitch angle to match the interactive stuff... most annoying...
Answers a question from yesterday, I guess. The expected answer, just not delivered in the way I desired...
So, once these are done, I rerun 8, 1 and 2 with zero pitch angle for validation, then redo the lot with a higher resolution. Currently running nine and have prepped one, leaving two to be done before doing the validation and the rest of it.
Right, 9 done, 1 going, 2 ready.
...
All nine done (yes, I spared you the details). Now have a full animation, in effect, of a deforming HBeta doppler profile, assuming a pitch angle of forty five degrees. Now set off the first of three validation runs for zero pitch angle. After that, will continue on and do the other six profiles at zero pitch angle. Once this is all done, I can grab a cosine distribution and use that. Have also used the original signal profiles and my own set of numbers to deduce correction factors to be introduced to try and get the relative intensities of the profiles correct. If I can then get one of the profiles and use it to calibrate its modelled counterpart, I'll see if I can actually model intensities through this method - assuming its any good at all... it may still not be.
...all to be done laterz...
...it is later. Will prepare the next eight runs for autorunning as background processes whilst I go home and sleep. Even with eight going, they should be done before I return tomorrow afternoon - unless someone logs on later in the hope of an empty Rimmer... it happens. Even so, that's their pigeon. After this has been done, I can start on the high res stuff. Exciting, really exciting...
Dammit - changing the pitch angle hasn't changed the autorun to the correct shape at all! Wtf is going on?
Aaaah - thermalisation level, I seeeeeee. There's a massive unshifted spike due to low level collisions. All to do with the efficiencies... sigh. Cross-sections for interactions, scattering ratios, emission efficiencies... there are so many atomic data to include in this damned thing. Still, can at least ramp up the thermalisation and give one set of results that can be used... also explains why the autoruns take so damned long...
ok, that's nine runs all happilly chugging along. One or two may even finish and get evaluated before I go... of course, I can tune the thermalisation all I want...
...time to go!
0 Comments:
Post a Comment
<< Home