This parameter controls the placement of the compression boxes within the input NDF (also see Parameter TRIM). It can take any of the following values:

• "ORIGIN" — The compression boxes are placed so that the origin of the pixel co-ordinate Frame  (i.e. pixel co-ordinates (0,0)) in the input NDF corresponds to a corner of a compression box. This results in the pixel origin being retain in the output NDF. For instance, if a pair of two-dimensional images which have previously been aligned in pixel co-ordinates are compressed, then using this option ensures that the compressed images will also be aligned in pixel co-ordinates.

• "FIRST" — The compression boxes are placed so that the first pixel in the input NDF (for instance, the bottom-left pixel in a two-dimensional image) corresponds to the first pixel in a compression box. This can result in the pixel origin being shifted by up to one compression box in the output image. Thus, images which were previously aligned in pixel co-ordinates may not be aligned after compression. You may want to use this option if you are using a very large box to reduce the number of dimensions in the data (for instance summing across the entire width of an image to produce a one-dimensional array).

• "LAST" — The compression boxes are placed so that the last pixel in the input NDF (for instance, the top-right pixel in a two-dimensional image) corresponds to the last pixel in a compression box. See the "FIRST" option above for further comments.

["ORIGIN"]

When there is an AXIS variance array present in the NDF and AXWEIGHT=TRUE the application forms weighted averages of the axis centres using the variance. For all other conditions the non-bad axis centres are given equal weight during the averaging to form the output axis centres. [FALSE]

Linear compression factors to be used to create the output NDF. There should be one for each dimension of the NDF. If fewer are supplied the last value in the list of compression factors is given to the remaining dimensions. Thus if a uniform compression is required in all dimensions, just one value need be entered. All values are constrained to be in the range one to the size of its corresponding dimension. The suggested default is the current value.

The NDF structure to be reduced in size.

When there are bad pixels present in the summation box these are ignored. Therefore a simple addition of the input-array component’s values will yield a result discordant with neighbouring output pixels that were formed from summation of all the pixels in the box. When NORMAL=TRUE the output values are normalised: the addition is multiplied by the ratio of the number of pixels in the box to the number of good pixels therein to arrive at the output value. When NORMAL=FALSE the output values are always just the sum of the good pixels. [TRUE]

NDF structure to contain compressed version of the input NDF.

If the input data type is to be preserved on output then this parameter should be set TRUE. However, this may result in overflows for integer types and hence additional bad values written to the output NDF. If this parameter is set FALSE then the output data type will be one of _REAL or _DOUBLE, depending on the input type. [FALSE]

Title for the output NDF structure. A null value (!) propagates the title from the input NDF to the output NDF. [!]

If Parameter TRIM is set TRUE, the output NDF only contains data for compression boxes which are entirely contained within the input NDF. Any pixels around the edge of the input NDF which are not contained within a compression box are ignored. If TRIM is set FALSE, the output NDF contains data for all compression boxes which have any overlap with the input NDF. All pixels outside the bounds of the NDF are assumed to be bad. That is, any boxes which extend beyond the bounds of the input NDF are padded with bad pixels. See also Parameter ALIGN. [current value]

If the input NDF contains bad pixels, then this parameter may be used to determine the number of good pixels which must be present within the addition box before a valid output pixel is generated. It can be used, for example, to prevent output pixels from being generated in regions where there are relatively few good pixels to contribute to the smoothed result.

WLIM specifies the minimum fraction of good pixels which must be present in the summation box in order to generate a good output pixel. If this specified minimum fraction of good input pixels is not present, then a bad output pixel will result, otherwise the output value will be the sum of the good values. The value of this parameter should lie between 0.0 and 1.0 (the actual number used will be rounded up if necessary to correspond to at least one pixel). [0.3]