Opened 14 years ago
Closed 12 years ago
#91 closed feature (fixed)
2D resolution calculations
Reported by: | srkline | Owned by: | srkline |
---|---|---|---|
Priority: | major | Milestone: | Wish List |
Component: | Analysis | Keywords: | |
Cc: | Blocking: | ||
Task: |
Description
As a first pass, use symmetric resolution (excluding gravity), then a 2D Gaussian can be convolved with the model using some sort of Igor built-in function.
Gravity would be more of a mess to add in, but there must be some way of casting it into a simpler problem...
Change History (6)
comment:1 Changed 14 years ago by srkline
comment:2 Changed 14 years ago by srkline
- Milestone Analysis 4.0 Release deleted
Important, but more work needs to be done than time permits before release. Table for now.
comment:3 Changed 14 years ago by srkline
- Milestone set to Analysis Wish List
comment:4 Changed 12 years ago by ajj
- Milestone changed from Analysis Wish List to Wish List
comment:5 Changed 12 years ago by srkline
mostly done. Use the full resolution function, including gravity. 10x10 quadrature is sufficient for the test cases I've run. I'm integrating in (parallel,perpendicular) coordinates at each point, rather than casting into XY. Takes about 10s for the whole detector on a slow machine (1GHz, 2 processor). Should be sufficiently fast on a modern computer.
needs just a little more testing and documentation
comment:6 Changed 12 years ago by srkline
- Resolution set to fixed
- Status changed from new to closed
I think we're done here. Testing and actual use will tell if there are any issues.
In principle, I've got the necessary equations from David. All that is to do is some sane implementation. Quadrature (even 5x5) makes the calculation 25x slower!!!. Try to look for an approximation that allows us to use an "Image..." Igor function, or write an XOP to do the calculation (messy Xfunc calling Xfunc).