Open ijiraq opened 4 years ago
Hi JJ,
the *.eff files can all be in the directory for the survey we want to simulate, at all time. What matters is the content of file pointings.list in the said directory. This is the one that defines the blocks or surveys we want to simulate. For each pointing, it tells the program which .eff file to use. So one can imagine having all .eff files in the directory, and having a script or program or web-interface that generates the pointings.list file from an ensemble of pointing lines that give the usual informations: time, pointing, size or shape of the FOV, efficiency file, filling factor, ...
By the way, I'm considering creating more pointing lines for other surveys. I've already done that for the Fuentes & Holman (2008) survey, and partly for some large scale surveys. Depending on the survey, it can be used to simply determine a population estimate for the full KBO population. I used that for the LF paper with FH08 survey. Obviously, in this case, having only the main belt in my model, I get an indication of order of magnitude, since the objects in FH08 were not tracked, hence have no dynamical classification. But there exist other large scale shallow surveys for which we should be able to get most of the objects tracked. Hence we can do a better job a simulating them, like we did for CFEPS, so this gives us another constraint, in particular for the large objects, with increased area coverage.
Cheers, Jean-Marc
What you say above is basically what I would advocate for also. Have one directory with just the .eff files. That is the data base of survey efficiencies. That can even have multiple sub-directories but the same .eff file should only ever appear once. Then there will be a pointing that is in a separate area and just indicates which .eff files (blocks of data) are to be used in a given simulation.
Of course, there are details in the pointing files now that are not in the .eff files (I think?) so we'd need to move some new information into the .eff files so that a point file can be generated off them.
Instead of dealing with the files in multiple directories we could store the files in a master directory and then have a script that linked the survey blocks the user is interested in into the appropriate directory for the fortran code to find it.