Hi folks
I am more than a little confused now (my usual state) !
It seemed to me that the initial task was to sum elements of a real array
such that a 'true' double precision result.
I must assume that sum was a surrogate for other arithmetic functions that
might be performed on all elements of a single array, or that the person
wanted the convenience of inputting fixed constant arrays to a lesser
precision, because .....
I recall that someone in the process has referred to basic error-propagation
theory - namely that the absolute errors are summed for addition and
subtraction, and the proportional errors summed for muliplication and
division (etc).
In that event, (for individual array elements of approximately the same
proportional error - ie. a significant number of significant figures) the
proportional error of the result can be no better (in fact approximately the
same) as the input.
On another aspect, it seems perfectly logically consistent with current
compiler behavoiur if, when given 'cart blanche', it were to sum the
elements as single precision and then convert the result to double
precision.
Clearly I have missed the point. I would be grateful for a brief personal
note from someone to confirm my simplistic nature and the error of my ways.
(pun intended but NOT irony)
Gary H
************************************************************************
The information in this e-mail together with any attachments is
intended only for the person or entity to which it is addressed
and may contain confidential and/or privileged material.
Any form of review, disclosure, modification, distribution
and/or publication of this e-mail message is prohibited.
If you have received this message in error, you are asked to
inform the sender as quickly as possible and delete this message
and any copies of this message from your computer and/or your
computer system network.
************************************************************************
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|