Clean is an iterative algorithm based on the assumption that the image can be well represented by a superposition of point sources.  Often it is satisfactory for extended sources as well.

The Clean algorithm starts with a dirty map created by the back projection algorithm.  Clean then tries to determine what set of point sources could have produced the dirty map given the instrument point spread function (PSF). The PSF is the dirty map created by a point source at a given pixel location, i.e. the imager's response to a delta function source. 

The Clean algorithm first finds the pixel with the highest flux in the dirty map. It assigns a point source with a fixed fraction (known as the “loop gain”) of that flux at that pixel location in a new map of Clean components. It then subtracts that same fractional flux from the dirty map but spread out according to the PSF centered on that pixel. This process is repeated by taking the pixel with highest flux in the new dirty map, the so-called residual map. Cycling through this process can be continued a specified number of times or until the peak absolute flux in the residual map is negative.

The Clean algorithm uses the Cartesian or ANNSEC coordinate system internally.

Using the Clean algorithm at the command line:

Select algorithm 

o->set, image_algorithm='clean'

or 'cl' or 'hsi_clean'

Object Class 


Extract Object

clean_obj = o->get(/obj,class='hsi_clean') 

Extract the object used in the hsi_image object, o

Parameter Prefix 

clean, e.g.
clean_chi_sq_tot = o->get(/clean_chi_sq_tot)
clean_control = o->get(/clean, /control)
clean_info = o-> get(/clean, /info)
clean_all = o->get(/clean)

All parameter names specific to this algorithm have this prefix

Parameter Table 

Clean Object Parameter Table

List and short description of control and info parameters specific to this algorithm