Is IEEE 754 floating point a good idea to use NN (non-A-number) which is non-mathematical Reasons?
In our case, they have not been set up yet because the price has not been obtained from any other device. Reference IEC1131 is an embedded system using REAL32 values. Edit: The programming language is C, so we will most likely use NAN and isnanf (x), which are from C99. However we need some additional contortions to bring them into the OS compatibility layer Might be possible.
In the programming languages, the default seems to begin with floating point variables with zero, whose internal representation is all zero is not useful for us, because 0 is in the range of valid values.
It seems that there is a clean solution to use NaN, but maybe this is more trouble, and we should choose something
Just paid attention to this question.
This is one of the uses of NAN that is the attention of the IEEE 754 Committee (I was a member of a committee) In the arithmetic, the NN publicity rules make very attractive because if you have some initial data There is a long sequence of calculations involving, so you will not mistake the result for a valid result. It can be put back through your calculations so that you can find out how fast you use the initial data.
He said, there are some disadvantages which are beyond the control of the 754 committee: As other people have noted, all hardware does not support the speed of Nayn, which is at risk of performance May be needed. Fortunately, one does not have too many operations on initial data in display-critical settings often.
Comments
Post a Comment