search

H-E-L-P / XID_plus

Next generation XID development
MIT License
3 stars 8 forks source link

PRF problem #2

Closed pdh21 closed 9 years ago

pdh21 commented 10 years ago

At the moment it is expecting the Gaussian2Dkernel from astropy. Should be an array (with odd number of rows/columns)

pdh21 commented 10 years ago

I definitely need to change this now as looks like there might be an issue with Gaussian psf

pdh21 commented 9 years ago

Investigating PRF

Looking at the COSMOS test region fit, I stacked on the residual map and found the following: image

Stacking on the original map, gives the expected PRF: image

It appears as if the PSF is too wide, or too much power is being given to slopes of PSF. I guess this is to be expected if the Gaussian is made wider to compensate for power at wings.

Solution:

I am going to try using the Corrected SPIRE PSFs from: http://www.astro.princeton.edu/~ganiano/Kernels/Ker_2012_May/PSF_fits_Files/

This will require change to code to allow arrays rather than astropy's kernel.