Appendix J
Configuration Parameters Listed by Category

 J.1 General
 J.2 Diagnostics
 J.3 Pre-processing
 J.4 Iterative: COM model
 J.5 Iterative: NOI model
 J.6 Iterative: FLT model
 J.7 Iterative: EXT model
 J.8 Iterative: AST model

This appendix orders the configuration parameters described in Appendix I into functional categories.

J.1 General





Parameter

Purpose

Values







numiter

Specifies when to stop iterating.

default : -5
dimmconfig_jsa_generic : -25
dimmconfig_bright_extended : -40
dimmconfig_bright_compact : -40
dimmconfig_blank_field : 4




maptol

Specifies when to stop iterating.

default : 0.05
dimmconfig_jsa_generic : 0.01




modelorder

Specifies which models to include in the iterative process, and the order in which they are evaluated.

default : (com,gai,ext,flt,ast,noi)
dimmconfig_blank_field : (com,ext,ast,noi)




hitslimit

Rejects map pixels that receive very few samples.

default : 0.01





J.2 Diagnostics





Parameter

Purpose

Values







fakemap

Diagnostic tool to explore the effects of the map-making process on known sources.

default : <undef>




fakescale

Control the use of the supplied fake map.

default : 1




exportndf

Create NDFs holding final model values.

default : 0




itermap

Create NDFs holding the map created by each iteration.

default : 0




bolomap

Create NDFs holding the map made form each bolometer.

default : 0




shortmap

Create NDFs holding the map made form a short chunk of data.

default : 0




diag.out

Switches on the dumping of various diagnostic information.

default : <undef>





J.3 Pre-processing





Parameter

Purpose

Values







downsampscale

Speeds up map-making, and reduces memory requirements.

default : -1




maxlen

Determines how the input time-series data is split into chunks

default : 0




doclean

Allows pre-cleaned data to be used.

default : 1




exportclean

Allows the initial cleaned data to examined or saved for later use.

default : 0




order

Baseline removal

default : 1




badfrac

Ensures that bad data from DA system are ignored

default : 0.05




compreprocess

Remove common-mode before the iterative algorithm begins.

default : 0




dcthresh

Control the cleaning of DC steps in bolometer time streams

default : 25.0
dimmconfig_jsa_generic : 100
dimmconfig_bright_compact : 100




dcfitbox

Control the cleaning of DC steps in bolometer time streams

default : 30




dcmaxsteps

Control the cleaning of DC steps in bolometer time streams

default : 4




dclimcorr

Control the cleaning of DC steps in bolometer time streams

default : 0




dcsmooth

Control the cleaning of DC steps in bolometer time streams

default : 50




spikethresh

Controls time-based spike detection within initial data cleaning.

default : 0
dimmconfig_blank_field : 10




spikebox

Controls time-based spike detection within initial data cleaning.

default : 50




whiten

Experimental.

default : 0




noisecliphigh

Reject bolometers based on their noise.

default : 4
dimmconfig_jsa_generic : 10.0
dimmconfig_bright_compact : 10.0




noisecliplow

Reject bolometers based on their noise.

default : 0




noiseclipprecom

Reject bolometers based on their noise.

default : 0




flagslow

Flag data when we’re moving too slowly

default : 30




flagfast

Flag data when we’re moving too fast

default : 1000




filt_edgelow

Specifies the highest frequency to be retained by the initial data cleaning.

default : 0




filt_edge_largescale

Specifies the largest scale size to be retained by the initial data cleaning.

default : 0
dimmconfig_blank_field : 200




filt_edge_smallscale

Specifies the smallest largest scale size to be retained by the initial data cleaning.

default : 0




filt_order

Indicates the shape of the filter.

default : 0





J.4 Iterative: COM model





Parameter

Purpose

Values







com.perarray

Controls the estimate of the common-mode signal

default : 0
dimmconfig_jsa_generic : 1
dimmconfig_bright_compact : 1
dimmconfig_blank_field : 1




com.sig_limit

Controls the rejection of time slices with inconsistent common-modes.

default : 0




com.noflag

Controls the rejection of bad samples from the COM estimate

default : 0




com.corr_tol

Controls the rejection of bad samples from the COM estimate

default : 5
dimmconfig_bright_compact : 7




com.corr_abstol

Controls the rejection of bad samples from the COM estimate

default : 0.2




com.freeze_flags

Controls flagging of samples that differ from the common-mode.

default : 0




com.gain_tol

Controls the rejection of bad samples from the COM estimate

default : 5
dimmconfig_bright_compact : 7




com.gain_abstol

Controls the rejection of bad samples from the COM estimate

default : 3
dimmconfig_bright_compact : 5




com.gain_box

Controls the rejection of bad samples from the COM estimate

default : -30.0




com.gain_fgood

Controls the rejection of bad samples from the COM estimate

default : 0.25




com.gain_rat

Controls the rejection of bad samples from the COM estimate

default : 4




com.zero_mask

Provides a better estimate of the common-mode ("COM") signal, by excluding samples that fall within fixed regions on the sky specified by an external mask.

default : 0




com.zero_circle

Improves common-mode estimation by excluding sources within a circle of given radius from the COM estimate.

default : <undef>




com.zero_lowhits

Improves common-mode estimation by excluding sources in regions containing many data samples.

default : 0




com.zero_snr

Improve the estimate of the common-mode by excluding samples that correspond to high SNR pixels in the map.

default : 0




com.zero_snr_ffclean

Provides alternative method for SNR masking.

default : 0




com.zero_snrlo

Improve estimate of the common-mode by increasing the size of the SNR mask without introducing noise.

default : <undef>




com.zero_union

Controls how multiple COM masks are combined.

default : 1




com.zero_freeze

Prevent the COM mask from changing after a given number of iterations. This can help convergence.

default : 0.0





J.5 Iterative: NOI model





Parameter

Purpose

Values







noi.calcfirst

Determines when the noise in each bolometer is estimated.

default : 0




noi.box_size

Allow finer estimation of the noise levels in the time-series data.

default : -15




noi.box_type

Determines how the noise in each box is found.

default : 1





J.6 Iterative: FLT model





Parameter

Purpose

Values







flt.filt_edge_largescale

Specifies the largest scale size to be retained by the FLT model.

default : 
dimmconfig_jsa_generic : 200
dimmconfig_bright_extended : 480
dimmconfig_bright_compact : 200




flt.filt_edge_largescale_last

Specifies the largest scale size to be retained by the FLT model on the last iteration.

default : <undef>




flt.filt_order

Indicates the shape of the filter.

default : 0




flt.ring_box1

Controls the flagging of samples that suffer from ringing.

default : 0




flt.zero_mask

Speeds up convergences and reduces ringing by excluding sources within a region specified by an external mask file from the filtering performed by the FLT model.

default : 0




flt.zero_circle

Speeds up convergences and reduces ringing by excluding sources within a circle of given radius from the FLT estimate.

default : <undef>
dimmconfig_bright_compact : 0.016666




flt.zero_lowhits

Experimental.

default : 0




flt.zero_snr

Speeds up convergences and reduces ringing by excluding samples that correspond to high SNR pixels in the map.

default : 0
dimmconfig_jsa_generic : 5
dimmconfig_bright_extended : 5




flt.zero_snr_ffclean

Provides alternative method for SNR masking.

default : 0




flt.zero_snrlo

Speeds up convergences and reduces ringing by increasing the size of the SNR mask without introducing noise.

default : <undef>
dimmconfig_jsa_generic : 3
dimmconfig_bright_extended : 3




flt.zero_niter

Allows FLT masking to be switched off after a given number of iterations.

default : 2




flt.zero_union

Controls how multiple FLT masks are combined.

default : 1




flt.zero_freeze

Prevent the FLT mask from changing after a given number of iterations. This can help convergence.

default : 0.0




flt.notfirst

May improve convergence by avoiding the filtering of strong sources on the first iteration.

default : 0





J.7 Iterative: EXT model





Parameter

Purpose

Values







ext.tausrc

Controls the extinction values used in the EXT model.

default : auto




ext.taumethod

Controls the extinction values used in the EXT model.

default : adaptive




ext.taurelation.450

Controls the 450 um extinction values used in the EXT model and the EXTINCTION task.

default : (23.3,-0.018,0.05)




ext.taurelation.850

Controls the 850 um extinction values used in the EXT model and the EXTINCTION task.

default : (3.71,-0.040,0.202)




ext.csotau

Controls the extinction values used in the EXT model.

default : <undef>




ext.filtertau

Controls the extinction values used in the EXT model.

default : <undef>





J.8 Iterative: AST model





Parameter

Purpose

Values







ast.mapspike

Removes spikes from the map.

default : 10




ast.skip

Skip subtraction of astronomical signal.

default : 0
dimmconfig_jsa_generic : 5
dimmconfig_bright_extended : 5




ast.zero_mask

Reduces spurious large scale structure in the final map within fixed regions specified by an external mask.

default : 0




ast.zero_circle

Reduces spurious large scale structure in the final map outside a circle of given radius.

default : <undef>
dimmconfig_bright_compact : 0.016666




ast.zero_lowhits

Reduces spurious large scale structure in the final map in regions containing few data samples.

default : 0




ast.zero_snr

Reduces spurious large scale structure in the final map within regions of low signal-to-noise.

default : 0.0
dimmconfig_jsa_generic : 5
dimmconfig_bright_extended : 3




ast.zero_snr_ffclean

Provides alternative method for SNR masking.

default : 0




ast.zero_snrlo

Can help to remove bowls around sources by increasing the size of the SNR mask without introducing noise.

default : <undef>
dimmconfig_jsa_generic : 3
dimmconfig_bright_extended : 2




ast.zero_snr_fwhm

Can help to remove bowls around sources.

default : 0




ast.zero_snr_low

Can help to remove bowls around sources.

default : -1.1




ast.zero_union

Controls how multiple AST masks are combined.

default : 1




ast.zero_freeze

Prevent the AST mask from changing after a given number of iterations. This can help convergence.

default : 0.0




ast.zero_niter

Allows AST masking to be switched off after a given number of iterations.

default : 0




ast.zero_notlast

Allows flux to be present in masked areas in the final map.

default : 1