next up previous contents index
Next:  CHANGE SCALE Up: Fit Sub-Menu Previous:  Advice on Model Fitting


 The Data Mask

Allowing elements to be arbitrarily included or excluded from fitting and other operations is a very powerful facility which allows much flexibility in the data analysis process. ``Bad'' data points can be excluded from operations. Optimum non-rectangular data regions can be selected reducing calculations for fitting and other calculation intensive operations.

The ``mask'' is a 2-D area of the same size as the current data and memory arrays. Each data element may be ``masked'' or not. By default no elements are masked. The command DEFINE MASK (See Section 17.8, Page [*]) allows masked-off regions to be defined, and the command INPUT PARAMETERS (See Section 17.11, Page [*]) also calls the DEFINE MASK routine. The command CLEAR MASK (See Section 17.5, Page [*]) resets all elements to ``unmasked''.

Alternatively the ``masked-off'' areas can be defined from the data values themselves by using the THRESHOLD MASKING command. This allows a threshold value to be defined and all data elements with values above, or below (as required), this value are defined as ``masked-off'' (see Section 17.25, Page [*]). The mask can be ``transferred'' to the memory for further use and possible storage using the TRANSFER MASK TO MEMORY command (see Section 17.27, Page [*]).


next up previous contents index
Next:  CHANGE SCALE Up: Fit Sub-Menu Previous:  Advice on Model Fitting
Andrew Hammersley
2004-01-09