STScI Logo

msstatistics stsdas.toolbox.imgtools.mstools


NAME · USAGE_ · DESCRIPTION · PARAMETERS · PSETS · EXAMPLES
BUGS · REFERENCES · SEE_ALSO

NAME

msstatistics -- statistics in IMSETs.

USAGE

msstatistics input

DESCRIPTION

This task is an extension of the "gstatistics" task. It supports OIF, GEIS, STIS and NICMOS file formats.

It computes and prints in tabular form a number of statistical quantities, either for each individual group/IMSET in each input file or for an accumulation over a range of groups/IMSETs in each input file. Results for the last file/group/IMSET are saved into parameters in the `egstp' pset. Each individual HDU in multi-IMSET files (as in STIS and NICMOS) is handled separately from the others.

The task accepts mixed file types into a single input list. To discriminate in between the several supported types it uses primarily the file name extension. OIF images have ".imh" name extensions. GEIS files have name extensions of the form ".??h". STIS and NICMOS files have ".fits" or ".fit" name extensions. The task also looks into some header keywords to figure out the file format details. In GEIS files it reads in the GCOUNT keyword. In FITS files it looks for the INSTRUME and NEXTEND keywords. If INSTRUME has neither the value of STIS nor NICMOS, the FITS file is handled as an OIF image, thus only the first FITS extension is processed.

The task accepts Data Quality (DQ) information taken from appropriate sources. OIF images and GEIS files may have their DQ information stored in other files which have the same name as the input file but with a different file name extension. Both regular image files and IRAF ".pl" pixel lists may be used. Alternatively, an optional list/template of DQF files might be used. That list may either have the same number of elements as the main input list, or one single DQ file that will be applied to all input files. The files in the DQ list/template supersede any DQFs already associated to the input files by the name extension mechanism. The above do not apply to STIS and NICMOS images, since they always carry internally their DQ information.

The task has instrument-dependent psets that enable the user to select which DQ bits are actually used to mask out individual pixels from the stat computations. To provide more flexibility and independency from instrument-specific bit assignations, Data Quality bits can also be specified in generic form by using the `dqbits' pset. The actual boolean mask used to flag pixels out from the stat computations is built from both the generic bit pset and instrument-specific pset.

The task also accepts an optional list/template of mask files. These files may have any name, but the "masks" list must be paired with the primary "input" input list. Alternatively, one single mask can be applied to all files in the input list. These masks are intended to be used primarily as "science" masks, to remove objects/regions from the stat computations. Both regular images and pixel lists are acceptable. Contrary to DQ files, however, no bit-level information is taken from these masks; any pixel with a value different from zero will tell the task to discard the corresponding image pixel. No multi-group or multi-IMSET mask images are supported in this version.

Image sections are fully supported.

In the case of STIS and NICMOS files, it is up to the user to supply to the task a properly formatted file, in the sense that its NEXTEND keyword agrees with the number of FITS extensions, and the EXTVER number sequence goes from 1 up to the number of IMSETs in the file. Note however that the extensions themselves do not have to be physically stored in the file in any pre-determined sequence. FITS files which have a primary header and extensions but that do not agree with either NICMOS or STIS formats, are considered corrupted and may cause erratic behavior. This usually takes the form of either: (i) the task prints a header at STDOUT but no stats are computed and printed, or (ii) the task prints error messages from IRAF's FITS kernel about extensions not being specified. Files like these can be input to this task by supplying the extension number enclosed in brackets as part of the file name.

It's worth saying that msstatistics will not compute statistics for the Data Quality arrays in STIS and NICMOS files, since these arrays do not hold numerical measurement values, but bit-encoded flags for which statistical estimators are meaningless.

The parameter `groups' may be used to specify a range of groups in GEIS files or a range of IMSETS in a multi-extension FITS file (as in STIS and NICMOS files). The default value of `groups' is set to "*" indicating statistics will be done for all groups/IMSETS in each file. A group/IMSET range may be specified by a set of one to three integers. The general format is "g1-g2xg3", where g1 is the starting group/IMSET, g2 the ending group/IMSET and g3 the increment. For instance, 3-11x2 means a range of groups/IMSETs starting from group/IMSET 3 through group/IMSET 11 with a step of 2. If g3 is omitted it defaults to 1. A single group/IMSET g1 may be specified too. The group/IMSET range list is overridden if the input file(s) itself is a GEIS file and already includes a group specification in its name.

Statistics to be printed are those selected by the `stats' parameter, or they may all be selected by setting `stats="doall"'. The user may choose any combination of the following:

	 doall - all of the statistical quantities
	  npix - number of valid pixels
	   min - minimum pixel value
	   max - maximum pixel value
	   sum - sum over all the pixels
	  mean - mean of the pixel distribution
	stddev - standard deviation of the pixel distribution
	 midpt - estimate of the median of the pixel distribution
	  mode - estimate of the mode of the pixel distribution
	  skew - skewness of the pixel distribution
          kurt - kurtosis of the pixel distribution
         wmean - weighted mean of the pixel distribution
          wvar - weighted variance of the pixel distribution

Statistical quantities are computed using the expressions listed below.

       mean = sum (x1,...,xN) / N
     dev[i] = x[i] - mean
   variance = sum {dev[i]**2} / (N-1)
     stddev = sqrt (variance)
   skewness = sum {(dev[i] / stddev) ** 3} / N
   kurtosis = [sum {(dev[i] / stddev) ** 4} / N] - 3
      wmean = sum{x[i]/(s[i]*s[i])} / sum{1/(s[i]*s[i])}
       wvar =  1 / sum{1/(s[i]*s[i])}

where s[i] is the i-th pixel's error, available only in STIS and NICMOS IMSETs. The weighted quantities are computed only for the science array.

The task computes all specified statistical quantities either separately for individual groups/IMSETs of an image when `gaccum' is set to "no", or accumulatively over all groups/IMSETs specified by the parameter `groups' when `gaccum' is set to "yes". The values written to the `egstp' pset are either the last file/group/IMSET's, or the ones accumulated across all groups/IMSETS in the last file.

The median is estimated by integrating the data histogram and computing by interpolation the data value at which half the pixels are below that data value and half are above it. The mode is estimated by locating the maximum of the data histogram and fitting the peak by parabolic interpolation. While a histogram with the number of bins up to one fourth of the number of pixels is used to achieve high accuracy, re-binning the histogram to a relatively low resolution may be beneficial for searching the peak of the histogram. Successive histograms with bin widths twice as broad as the previous one are used to obtain intermediate results of the mode. The re-binning of the histograms is ended when either the bin width is greater than 0.01 * standard deviation or the number of bins becomes less than one tenth of the number of pixels. The final result for the mode is an average of all the intermediate results of the mode.

The `nsstatpar' and `wfdqpar' psets provide the necessary instrument and format-dependent information. The `wfdqpar' pset holds the Data Quality File name extension for GEIS files and OIF images. This is usually ".c1h" for WFPC images, but this pset can be used to specify the DQF name extension for any other GEIS format file, and for OIF images as well. The special case of pixel lists is handled by setting `dqfext = ".pl"'. The remaining parameters hold the individual flags for each Data Quality bit in WFPC files.

The `nsstatpar' pset is specific for STIS and NICMOS. The `arrays' parameter is used to specify which HDUs in each IMSET in each file will have their stats computed. The `clarray' parameter specifies which HDUs in the last file/IMSET will have its stats written to the `egstp' pset. The `stisdqpar' and `nicdqpar' psets hold the individual flags for each Data Quality bit in respectively STIS and NICMOS files.

PARAMETERS

input [file name list/template]
List of files for which pixel statistics are to be computed.
(masks = "") [file name list/template]
List of masks which can be either pixel list files or mask images.
(dqf = "") [file name list/template]
List of Data Quality Files which supersede any DQFs associated by file name extension (OIF/GEIS formats only).
(groups = "*") [string]
Range list of groups/IMSETs to be processed (syntax is g1-g2xg3).
(stats = "npix,mean,stddev") [string]
The statistical quantities to be computed and printed.
(lower = INDEF) [real]
Use only pixels with values greater than or equal to this limit. All pixels are above the default value of INDEF.
(upper = INDEF) [real]
Use only pixels with values less than or equal to this limit. All pixels are below the default value of INDEF.
(gaccum = no) [boolean]
Accumulate statistics across groups/IMSETs ?
(egstp = [pset])
Pset name for storing the resulting statistical parameters.
(namecol = 25) [int, min=1, max=80]
Column width used by file name in output printout.
(dqon = yes) [boolean]
Use Data Quality information to mask pixels out from the stat computations ? If not explicitly set, no Data Quality information will be used by the task.
(dqbits = "") [pset]
Data Quality bits to look for in generic files. The flags set in this pset will be logically ORed with the flags set in the instrument-specific pset.
(nsstatpar = [pset])
Pset with STIS and NICMOS-dependent parameters.
(wfdqpar = [pset])
Pset with WFPC-dependent parameters. Also for GEIS files and OIF images.

PSETS

NSSTATPAR
(arrays = "science,error,time,sample") [string]
HDUs to be processed.
(clarray = "science") [string]
HDU whose stats are written into output pset.
(stisdqpar = "") [pset]
Data Quality bits to look for in STIS IMSETs.
(nicdqpar = "") [pset]
Data Quality bits to look for in NICMOS IMSETs.
WFDQPAR
(dqfext = ".c1h") [string]
Data Quality File name extension. Valid also for any GEIS file or OIF image.
(wfdqpar = "") [pset]
Data Quality bits to look for in WFPC files.

EXAMPLES

BUGS

Because of current limitations in handling INDEFs in C code, INDEF is not used by this task to signal that something cannot be computed. Zero is used instead, which makes the output listing a bit confusing, since genuine zero-valued results cannot be told apart from indefinite-valued results.

REFERENCES

This task was written by I. Busko.

SEE ALSO


Source Code · Package Help · Search Form · STSDAS