[gtkada] how to print a value with 16 bit precision

[Damien Carbonne]

nabh4u a écrit :
> Thank you Damien for the information. Is this another procedure i need to
> include in my program? can you tell me how can i add this part in my code?
> Thank you once again for the reply. 
>
>   
You don't "include" (There is no "include" in Ada, but this is another 
story) a procedure but a package into which this procedure is defined.
The package that contains this version of Put is already withed, it is 
Ada.Long_Float_Text_IO;
In fact you are already using it, but with default values for last 3 
parameters.
When you write:
(1)   Put (eps)
the compiler understands this:
(2)   Put (eps, Default_Fore, Default_Aft, Default_Exp);
Those default values are defined like this in Ada.Text_IO.Float_IO
   Default_Fore : Field := 2;
   Default_Aft  : Field := Num'Digits - 1;
   Default_Exp  : Field := 3;

Ada.Text_IO.Float_IO is a generic package which is instantiated with 
Long_Float to give Ada.Long_Float_IO.
So (1) is equivalent to:
   Put (eps, 2, 14, 3);


Practically, you need to change:
   Put(cnt + eps + eps + eps + eps + eps + eps);
into:
   Put(cnt + eps + eps + eps + eps + eps + eps, Aft => 16);

Do the same thing for the other uses of Put with a Long_Float (You can 
find them yourself !)

As indicated by Christoph Grein in his reply, this is indeed dubious as Float has 15 digits.
But this is what you (your teacher?) asked.

Regards

Damien




> Damien Carbonne wrote:
>   
>> Hi,
>>
>> You are simply using default parameters of Put here:
>>    procedure Put
>>      (Item : Num;
>>       Fore : Field := Default_Fore;
>>       Aft  : Field := Default_Aft;
>>       Exp  : Field := Default_Exp);
>>
>> and Default_Aft is defined like that :
>>    Default_Aft  : Field := Num'Digits - 1;
>>
>> As Long_Float has 15 digits, that explains the 14 printed ones.
>>
>> So you need to set Aft to the right value.
>> It could be like that:
>> Put (Foo) -> Put (Foo, Aft => 16);
>>
>> Damien
>>
>>
>> nabh4u a écrit :
>>     
>>> Hello everyone,
>>>
>>> Gurus i need help. I am new to Ada programming. It is very urgent, please
>>> help me. i have to submit this by tomorrow. I have a small project where
>>> i
>>> compute the machine epsilon and i have to print the value with a 16 bit
>>> precision. I have tried long_float and i am getting 14 bit precision but
>>> i
>>> want a 16 bit precision. i have also tried to use type but was not
>>> successful.
>>>
>>> my code is:
>>>
>>> with Ada.Text_IO,Ada.Integer_Text_IO,Ada.Long_Float_Text_IO;
>>> use Ada.Text_IO,Ada.Integer_Text_IO,Ada.Long_Float_Text_IO;
>>> Procedure Eps is
>>> eps:Long_Float := 1.0;
>>> epsp1:Long_Float;
>>> pass:Integer := 0;
>>> cnt:Long_Float := 1.0;
>>> Begin
>>>      New_Line(1);
>>>      Put("*** Computing Machine Epsilon using Ada Language ***");
>>>      New_Line(2);
>>>      epsp1 := eps + 1.0;
>>>      Put("       Pass ");
>>>      Put("     ");
>>>      Put(" Machine Epsilon Value ");
>>>      New_Line(2);
>>>      while(epsp1 > cnt) loop
>>>           pass := pass + 1;
>>>           eps := 0.5 * eps;
>>>           epsp1 := eps + 1.0;
>>>           Put(pass);
>>>           Put("      ");
>>>           Put(eps);
>>>           New_Line;
>>>      end loop;
>>>      New_Line;
>>>      Put("Final Eps Value : ");
>>>      Put(eps);
>>>      New_Line;
>>>      Put("The Value of 1+eps+eps+eps+eps+eps+eps is : ");
>>>      Put(cnt + eps + eps + eps + eps + eps + eps);
>>>      New_Line;
>>>      Put("The Value of 1+(eps+eps+eps+eps+eps+eps) is : ");
>>>      Put(cnt + (eps + eps + eps + eps + eps + eps));
>>>      New_Line(2);
>>> End Eps;
>>>
>>>   
>>>       
>> _______________________________________________
>> gtkada mailing list
>> gtkada at lists.adacore.com
>> http://lists.adacore.com/mailman/listinfo/gtkada
>>
>>
>>     
>
>