Digits of precision for floating point FITS values


This attribute gives the number of significant decimal digits to use when formatting floating point values for inclusion in the FITS header cards within a FitsChan.

By default, a positive value is used which results in no loss of information, assuming that the value is double precision. Usually, this causes no problems.

However, to adhere strictly to the recommendations of the FITS standard, the width of the formatted value (including sign, decimal point and exponent) ought not to be more than 20 characters. If you are concerned about this, you should set FitsDigits to a negative value, such as -15. In this case, the absolute value (+15) indicates the maximum number of significant digits to use, but the actual number used may be fewer than this to ensure that the FITS recommendations are satisfied. When using this approach, the resulting number of significant digits may depend on the value being formatted and on the presence of any sign, decimal point or exponent.

The value of this attribute is effective when FITS header cards are output, either using AST_FINDFITS or by the action of the FitsChan s sink routine when it is finally deleted.



All FitsChans have this attribute.