LEDAS  
Leicester Database and Archive Service
 
Mission
    Description
Products
    Databases
Raw Data
    Archive
Analysis Software
    Archive
Ginga Information
Ginga Gallery
Related Sites
Databases
GINGALAC
GINGABGD
GINGAFRF
GINGALOG
GINGAMODE
GINGARAW
View Ginga DBs
Line
Footer
e-mail

Ginga Specific Software: Program CLEANALL



Purpose

Various diagnostics are available to identify "bad" data in Ginga CUBEs and TIME series. For example, unpredictable increases in the background, particularly those due to electron flares, can cause the SUD and PI_MONI rates to depart from the normally linear relationship. For background data, a similar correlation exists between the SUD and LAC rates in good quality data. In addition, data "spikes" can be identified by comparing adjacent spectral channels in the LAC.

Program cleanall cleans any Ginga CUBE or TIME series by fitting a straight line to the SUD vs PI_MONI, SUD vs LAC and LAC vs LAC data. Points which lie more than CLN_NSIGMA standard deviations off this line are deleted.

This program can be used to clean both source and background observations, however if the source is expected to be variable, the SUD vs LAC correlation should be turned off (see below).


Input File

A Ginga CUBE or TIME series to be cleaned.


Output File

A "CLEANed" Ginga CUBE or TIME series.


Input Parameters

Cleanall requires four input parameters. The QX parameter system is used to read these parameters.

CLN_DATIN The name of the input file containing data to be cleaned.
CLN_DATOUT The name of the output file created, containing "CLEANed" data.
CLN_NSIGMA(3) The default values of CLN_NSIGMA are:
  1. CLN_NSIGMA(1)=3.0 (SUD vs PI_MONI)
  2. CLN_NSIGMA(2)=4.0 (SUD vs LAC)
  3. CLN_NSIGMA(3)=5.0 (LAC vs LAC)
Data points which lie more than CLN_NSIGMA standard deviations from the best-fit correlation are deleted. The above values have been chosen such that the majority of "bad" data will be detected with a minimum of "good" data being deleted by mistake.
Reducing the value of CLN_NSIGMA increases the sensitivity of cleanall to "bad" data, but also increases the amount of "good" data erroneously deleted.
For each correlation, cleaning can be turned off by setting the appropriate value of CLN_NSIGMA to zero. Therefore, for variable source data, CLN_NSIGMA(2)=0.0.
CLN_DUMP Controls whether the fit parameters derived in the correlation are written to the screen. The default is No, in which case only the number of points deleted in each correlation will be displayed.


Output Parameter

Cleanall returns one output parameter. The QX parameter system is used to display this parameters.

CLN_DATOUT The name of the output file containing "CLEANed" data. If no data points were deleted then CLN_DATOUT returns the name of the input file.


[Back to Contents]
 
HOME
SEARCH
SERVICES INFO SOFTWARE ViZieR BLASTA DSS
ARCHIVES ASCA CHANDRA GINGA ROSAT ARNIE