Re: double is integer?

Same problem.

In general, it is impossible to deal with the notion that a floating-point value is an
"exact" anything, integer or anything else.

double d= 10.0;
double q = d / 3.0;
double t = q * 3.0;

It is very unlikely that t == d, it is equally unlikely that
long int i = (long int) d;
if ((double) i == t)

will produce anything useful. For example, i will be 10, but (double)i will be 10.0, and
t is still 9.9999999999999. Also, why 'long int' instead of just 'long' (the 'int' is a
noise word)

It is reasonably safe to assume that any value you have is the result of some computation.

Note that if you don't want users to type in a value that is not an integer, you have to
limit what they can type. Or you have to analyze the text of what they type, not the
floating-point representation of what they typed.

Floating point arithmetic has a loose grasp of the concept of precision. The fact that
its looseness can be rigorously characterized does not make it any less loose.

Any attempts to determine if a floating point value is an exact integer are probably
doomed.

For one thing, just the idea of using long is suspect. Is not 5E20 a valid integer? Yes,
but it can't be represented in a long, so any transformation that passes through a long is
doomed to failure. Note that floating point can represent values that are larger than
LONGLONG (__int64). In effect, there is absolutely no way to perform this test reliably.

(If you can accept approximations, then you can format the floating point number and see
if there is anything after the decimal point. It still won't work right, but it will be
close to right more of the time than comparing the floating point representations. But it
also will fail for all the above reasons, but if you use, say, 3 digits of decimal
precision you will only fail some of the time, rather than pretty consistently all the
time. And you'll have to deal with the failures, so you haven't solved anything, really)
joe


On Mon, 5 Feb 2007 16:55:23 -0000, "GT" <ContactGT_remove_@xxxxxxxxxxx> wrote:

"Joseph M. Newcomer" <newcomer@xxxxxxxxxxxx> wrote in message
news:0dcas2lrmirbjg293b14fd9odq39ide0au@xxxxxxxxxx
This is problematic at best. While it might work, in general it is not a
safe thing to do
because what if your value is 2.0000000000001 or 1.99999999999999? After
doing a lot of
floating point computations, you may have a value close to but not
precisely an integer.

For example
double d = 10.0;
double q = d / 3.0;
double t = q * 3.0;

you might assume that d == t, which will not be true, and isInteger(d) may
be true but
isInteger(t) will not be.

On the whole, this is a very risky thing to consider doing. You should
never assume that
any computation involving a double will ever produce an integer value.
For that matter,
if the integer value is very large, the double might lose low-order
precision anyway.

There is no reliable way to actually compare doubles to integers and know
that the double
is an integer value. Typically, you might consider

if( (d - ceil(d)) < fuzzfactor)

but that is about as good as it gets.
joe


How about:

bool isInteger(const double& d)
{

long int i = (long int) d;
if ((double) i == d)
{
// we have an integer
return true;
}
else
return false;
}

Joseph M. Newcomer [MVP]
email: newcomer@xxxxxxxxxxxx
Web: http://www.flounder.com
MVP Tips: http://www.flounder.com/mvp_tips.htm
.

Relevant Pages

  • Re: Rounding of the double
    ... John von Neumann suggested that only fixed-point integers should be used because floating ... If you choose your representation as double, then you have to live with the consequences. ... you can never get precision in any computation on a computer than involves ... MVP Tips:http://www.flounder.com/mvp_tips.htm- Hide quoted text - ...
    (microsoft.public.vc.mfc)
  • Re: Subplot frustration
    ... an artifice of ML's representation of the internal handle as a default Double... ... converting to decimal representations of floating point numbers. ... to a decimal string with no loss of precision. ... My other supposition is that the "handle graphics" package was a product not developed inside TMW, but that again is an assumption. ...
    (comp.soft-sys.matlab)
  • Re: Will too many paradigms addle my brain.
    ... i.e. that a subclass should define a subtype. ... it's represented as an integer or as a floating point object. ... Most of the time I don't want to specify the precision explicitly, ... representation, and exact infinities are primarily used to provide ...
    (comp.lang.functional)
  • Re: Linear Algebra Challenge
    ... Since I'm using floating point, so I'll never be able to calculate one ... floating point math set to 99 digits. ... As close as I'm willing to wait if I use arbitrary precision. ... This mode is fast; when you select arbitry ...
    (comp.sys.hp48)
  • Re: Rounding off double precision
    ... operation has a small quantity of error built in. ... I understand how floating points are used to ... If I thought they were exact, ... because I didn't declare Pi to be double precision, ...
    (comp.lang.fortran)