The relative error of the values held in the took-up table


This attribute holds the relative error of the values held in the took-up table. It is used when simplifying a LutMap, to determine if the LutMap should be considered linear. Setting a larger value makes it more likely that a LutMap will be replaced by a WinMap (i.e. a linear Mapping) when simplified.

The default value is the value of the system constant DBL_EPSILON (typically around 1e-16 or 2E-16). If the values in the look-up table were derived from single precision data, it may be appropriate to set this attribute to a value around 1E-7.

Note, the value of this attribute may changed only if the LutMap has no more than one reference. That is, an error is reported if the LutMap has been cloned, either by including it within another object such as a CmpMap or FrameSet or by calling the AST_CLONE function.

Double precision.


All LutMaps have this attribute.