DRAWSIG - Draws $\pm$n standard-deviation lines on a line plot

Description:

This routine draws straight lines on an existing plot stored in the graphics database, such as produced by LINPLOT or HISTOGRAM. The lines are located at arbitrary multiples of the standard deviation (NSIGMA) either side of the mean of a given dataset. The default dataset is the one used to draw the existing plot. You can plot the lines horizontally or vertically as appropriate. The lines extend the full width or height of the plot's data area. Up to five different multiples of the standard deviation may be presented in this fashion. Each line can be drawn with a different style (see Parameter STYLE).

The application also computes statistics for those array values that lie between each pair of plotted lines. In other words it finds the statistics between clipping limits defined by each 2$*$NSIGMA range centred on the unclipped mean.

The task tabulates NSIGMA, the mean, the standard deviation, and the error in the mean after the application of each pair of clipping limits. For comparison purposes the first line of the table presents these values without clipping. The table is written at the normal reporting level.

Usage:

drawsig ndf nsigma [axis] [comp]

Parameters:

AXIS = LITERAL (Read)

The orientation of the lines, or put another way, the axis which represents data value. Thus the allowed values are "Horizontal", "Vertical", "X", or "Y". "Horizontal" is equivalent to "Y" and "Vertical" is a synonym for "X". On LINPLOT output AXIS would be "Y", but on a plot from HISTOGRAM it would be "X". The suggested default is the current value. ["Y"]

COMP = LITERAL (Read)

The name of the NDF  array component from which to derive the mean and standard deviation used to draw the lines: "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"]

DEVICE = DEVICE (Read)

The graphics device to draw the sigma lines on. [Current graphics device]

NDF = NDF (Read)

The NDF structure containing the data array whose error limits are to be plotted. Usually this parameter is not defined thereby causing the statistics to be derived from the dataset used to draw the plot. If, however, you had plotted a section of a dataset but wanted to plot the statistics from the whole dataset, you would specify the full dataset with Parameter NDF. [The dataset used to create the existing plot]

NSIGMA() = _REAL (Read)

Number of standard deviations about the mean at which the lines should be drawn. The null value or 0.0 causes a line to be drawn at the mean value.

STYLE = GROUP (Read)

A group of attribute settings describing the plotting style to use for the lines.

A comma-separated list of strings should be given in which each string is either an attribute setting, or the name of a text file preceded by an up-arrow character "$\wedge$". Such text files should contain further comma-separated lists which will be read and interpreted in the same manner. Attribute settings are applied in the order in which they occur within the list, with later settings overriding any earlier settings given for the same attribute.

Each individual attribute setting should be of the form:

$<$name$>$=$<$value$>$

where $<$name$>$ is the name of a plotting attribute, and $<$value$>$ is the value to assign to the attribute. Default values will be used for any unspecified attributes. All attributes will be defaulted if a null value (!) is supplied. See Plotting Attributes for a description of the available attributes. Any unrecognised attributes are ignored (no error is reported).

The attributes Colour(Curves), Width(Curves), etc., can be used to specify the style for the lines (Lines is recognised as a synonym for Curves). These values apply to all lines unless subsequent attributes override them. Attributes for individual clipping levels can be given by replacing Curves above by a string of the form "Nsig$<$i$>$" where "$<$i$>$" is an integer index into the list of clipping levels supplied for Parameter NSIGMA. Thus, "Colour(Nsig1)" will set the colour for the lines associated with the first clipping level, etc. The attribute settings can be restricted to one of the two lines by appending either a "$+$" or a "-" to the "Nsig$<$i$>$" string. Thus, "Width(Nsig2-)" sets the line width for the lower of the two lines associated with the second clipping level, and "Width(Nsig2+)" sets the width for the upper of the two lines. [current value]

Examples:

drawsig nsigma=3 style='style=1'

This draws solid horizontal lines on the last DATA picture on the current graphics device located at plus and minus 3 standard deviations about the mean. The statistics come from the data array used to draw the DATA picture.

drawsig phot 2.5

This draws horizontal plus and minus 2.5 standard-deviation lines about the mean for the data in the NDF called phot on the default graphics device.

drawsig phot 2.5 style='"colour(nsig1-)=red,colour(nsig1$+$)=green"'

As above, but the lower line is drawn in red and the upper line is drawn in green.

drawsig cluster [2,3] X Error

This draws vertical lines at plus and minus 2 and 3 standard deviations about the mean for the error data in the NDF called cluster on the default graphics device.

drawsig device=xwindows phot(20:119) 3 style="'colour=red,style=4'"

This draws red dotted horizontal lines on the xwindows device at $\pm$3 standard deviations using the 100 pixels in NDF phot(20:119).

Notes:

There must be an existing DATA picture stored within the graphics database for the chosen device. Lines will only be plotted within this picture.

Related Applications

KAPPA: HISTOGRAM, LINPLOT, MLINPLOT, STATS.

Implementation Status:

• This routine correctly processes the DATA, VARIANCE, and QUALITY, components of the NDF.

• Processing of bad pixels and automatic quality masking are supported.

• All non-complex numeric data types can be handled. The statistics are calculated using double-precision floating point.

• Any number of NDF dimensions is supported.