[squeak-dev] FloatArray gives unexpected results with float numbers.

Hello,

It is a bug.. or?
FloatArray newFrom: {0.938611557944869}. 
Expecting result is the entered float number, but evaluating gives - a FloatArray(0.938611567020416).

Nikolay

On Dec 20, 2008, at 5:54 PM, Nikolay Suslov wrote:

> Hello,
>
> It is a bug.. or?

No.

> FloatArray newFrom: {0.938611557944869}.
> Expecting result is the entered float number, but evaluating gives -  
> a FloatArray(0.938611567020416).

The comment in the #'newFrom:' methods reads as follows: "Answer an  
instance of me containing the same elements as aCollection." What the  
above does is take a float, add it to an Array, then create a  
FloatArray and add to it each element in the Array.

If you want just the float number, you already have it:

0.938611557944869 class == Float

James Foster

> Nikolay

On 21/12/2008, at 12:57 PM, James Foster wrote:

I think Nikolay's point is that the number isn't the same.

Antony Blakey
-------------
CTO, Linkuistics Pty Ltd
Ph: 0438 840 787

It's amazing that the one side of the conversation that survived is "I  
don't know art, but I know what I like". The reply from the artist was  
"Madam, so does a cow".
   -- Carl Kirkendall

On Dec 20, 2008, at 5:54 PM, Nikolay Suslov wrote:

> Hello,
>
> It is a bug.. or?

A misunderstanding.

> FloatArray newFrom: {0.938611557944869}.
> Expecting result is the entered float number, but evaluating gives -  
> a FloatArray(0.938611567020416).

Float has the following class comment: "My instances represent  
IEEE-754 floating-point double-precision numbers."
FloatArray has the following class comment: "FloatArrays store 32bit  
IEEE floating point numbers."

I think what you are seeing is the conversion of 64-bit floating point  
to 32-bit floating point. Sorry for the initial confusion about your  
expected result.

James

> Nikolay

2008/12/21 James Foster <[hidden email]>:
Even it that wouldn't be the case it would still be floating point
numbers and imprecise by design. Additionally not every decimal number
can be represented by a floating point number.

Cheers
Philippe

James,

Sorry for not clear problem statement,
Yes, the unexpected result is that the float number isn't the same, after adding to FloatArray.

Thanks,
Nikolay