The /NEGSCALED_NUMERIC qualifier enables RDML to convert the
    negative scaled numeric to floating (REAL or DOUBLE), depending
    upon the scale and data type, or to the original data type
    ignoring negative scale.

    Format:

            /NEGSCALED_NUMERIC = FLOATING     (Default)
            /NEGSCALED_NUMERIC = UNSCALED

    The default is FLOATING, which is the same as in previous
    versions.