STATS

Computes simple statistics for an NDF’s pixels

Description:

This application computes and displays simple statistics for the pixels in an NDF’s  data, quality or variance array. The statistics available are:

Iterative κ-sigma clipping may also be applied as an option (see Parameter CLIP).

Order statistics (median and percentiles) may optionally be derived and displayed (see Parameters ORDER and PERCENTILES). Although this can be a relatively slow operation on large arrays, unlike HISTAT the reported order statistics are accurate, not approximations, irrespective of the distribution of values being analysed.

Usage:

stats ndf [comp] [clip] [logfile]

Parameters:

CLIP( ) = _REAL (Read)
An optional one-dimensional array of clipping levels to be applied, expressed as standard deviations. If a null value is supplied for this parameter (the default), then no iterative clipping will take place and the statistics computed will include all the valid NDF pixels.

If an array of clipping levels is given, then the routine will first compute statistics using all the available pixels. It will then reject all those pixels whose values lie outside κ standard deviations of the mean (where κ is the first value supplied) and will then re-evaluate the statistics. This rejection iteration is repeated in turn for each value in the CLIP array. A maximum of five values may be supplied, all of which must be positive. [!]

COMP = LITERAL (Read)
The name of the NDF array component for which statistics are required: "Data", "Error", "Quality" or "Variance" (where "Error" is the alternative to "Variance" and causes the square root of the variance values to be taken before computing the statistics). If "Quality" is specified, then the quality values are treated as numerical values (in the range 0 to 255). ["Data"]
LOGFILE = FILENAME (Write)
A text file into which the results should be logged. If a null value is supplied (the default), then no logging of results will take place. [!]
NDF = NDF (Read)
The NDF data structure to be analysed.
ORDER = _LOGICAL (Read)
Whether or not to calculate order statistics. If set TRUE the median and optionally percentiles are determined and reported. [FALSE]
PERCENTILES( 100 ) = _REAL (Read)
A list of percentiles to be found. None are computed if this parameter is null (!). The percentiles must be in the range 0.0 to 100.0. This parameter is ignored unless ORDER is TRUE. [!]

Results Parameters

KURTOSIS = _DOUBLE (Write)
The population excess kurtosis of all the valid pixels in the NDF array. This is the normal kurtosis minus 3, such that a Gaussian distribution of values would generate an excess kurtosis of 0.
MAXCOORD( ) = _DOUBLE (Write)
A one-dimensional array of values giving the WCS co-ordinates of the centre of the (first) maximum-valued pixel found in the NDF array. The number of co-ordinates is equal to the number of NDF dimensions.
MAXIMUM = _DOUBLE (Write)
The maximum pixel value found in the NDF array.
MAXPOS( ) = _INTEGER (Write)
A one-dimensional array of pixel indices  identifying the (first) maximum-valued pixel found in the NDF array. The number of indices is equal to the number of NDF dimensions.
MAXWCS = LITERAL (Write)
The formatted WCS co-ordinates at the maximum pixel value. The individual axis values are comma separated.
MEAN = _DOUBLE (Write)
The mean value of all the valid pixels in the NDF array.
MEDIAN = _DOUBLE (Write)
The median value of all the valid pixels in the NDF array when ORDER is TRUE.
MINCOORD( ) = _DOUBLE (Write)
A one-dimensional array of values giving the data co-ordinates of the centre of the (first) minimum-valued pixel found in the NDF array. The number of co-ordinates is equal to the number of NDF dimensions.
MINIMUM = _DOUBLE (Write)
The minimum pixel value found in the NDF array.
MINPOS( ) = _INTEGER (Write)
A one-dimensional array of pixel indices identifying the (first) minimum-valued pixel found in the NDF array. The number of indices is equal to the number of NDF dimensions.
MINWCS = LITERAL (Write)
The formatted WCS co-ordinates at the minimum pixel value. The individual axis values are comma separated.
NUMBAD = _INTEGER (Write)
The number of pixels which were either not valid or were rejected from the statistics during iterative κ-sigma clipping.
NUMGOOD = _INTEGER (Write)
The number of NDF pixels which actually contributed to the computed statistics.
NUMPIX = _INTEGER (Write)
The total number of pixels in the NDF (both good and bad).
PERVAL() = _DOUBLE (Write)
The values of the percentiles of the good pixels in the NDF array. This parameter is only written when one or more percentiles have been requested.
SIGMA = _DOUBLE (Write)
The population standard deviation of the pixel values in the NDF array.
SKEWNESS = _DOUBLE (Write)
The population skewness of all the valid pixels in the NDF array.
TOTAL = _DOUBLE (Write)
The sum of the pixel values in the NDF array.

Examples:

stats image
Computes and displays simple statistics for the data array in the NDF called image.
stats image order percentiles=[25,75]
As the previous example but it also reports the median, 25 and 75 percentiles.
stats ndf=spectrum variance
Computes and displays simple statistics for the variance array in the NDF called spectrum.
stats spectrum error
Computes and displays statistics for the variance array in the NDF called spectrum, but takes the square root of the variance values before doing so.
stats halley logfile=stats.dat
Computes statistics for the data array in the NDF called halley, and writes the results to a logfile called stats.dat.
stats ngc1333 clip=[3.0,2.8,2.5]
Computes statistics for the data array in the NDF called ngc1333, applying three iterations of κ-sigma clipping. The statistics are first calculated for all the valid pixels in the data array. Those pixels with values lying more than 3.0 standard deviations from the mean are then rejected, and the statistics are re-computed. This process is then repeated twice more, rejecting pixel values lying more than 2.8 and 2.5 standard deviations from the mean. The final statistics are displayed.

Implementation Status:

Related Applications

KAPPA: HISTAT, NDFTRACE; FIGARO: ISTAT.