fprint Exponent printing question

mjamal · 10-20-2006, 10:43 AM

Hi,
I am writing a tool that needs to print doubles as exponential number with a certain precision and such.
I am using the %E modifier with the right precision (*.14), but my requirements need the number of digits after the E to be 3. Valid output example:
-9.00000000000000E-002
However, I can only get two digits after the E symbol... This is what I am doing right now:
double d = 1.232;
printf("double: %+.14E\n", d);
This prints as: +1.23200000000000E+00.

Thank you in advance for suggestions that may help!

Mansoor

exvor · 10-20-2006, 11:32 AM

Acording to "The C Programming Language" the E gives formatted output in this form ill retype what it says

Quote:

e,E double; [-]m.dddddde+/-xx or [-]m.ddddddE+/-xx, where the number of d's is given by the precision (default 6)

Thus I would surmize that printing output in the way that you need is not auto built into the printf function. You with either have to do 1 of 2 things.

1. Look around for a print function written in another library to do what you want.

2. Write a new printf (possibly from the printf source) that can output in the way that you want.

Good luck

This link might be helpful
http://www.mathworks.com/access/help...f/sprintf.html

firstfire · 10-22-2006, 12:24 AM

Hello!

As a kind of joke, you can use following:

Code:

#include <math.h>
#include <stdio.h>
#include <stdlib.h> // for `int abs(int)'
main()
    {
        double x = -1.1500018e18;
        int ex  = log10( fabs(x) );
        double mantissa = (x/pow(10.0, ex));
        printf("x=%+.14fe%c%.3d\n", mantissa, (ex<0)?'-':'+', abs(ex));
    }

You have to compile this with -lm flag:

Code:

gcc num.c -lm

WARNING: I strongly recommend you to not use this approach in scientific software!
It's: a) very inefficient, b) may cause overflows and loosing of accuracy, c)maybe smth. else.

Nevertheless, I hope, this is useful for you.

makyo · 10-22-2006, 02:27 PM

Hi.

Fortran seems more flexible, allowing you to specify the width of the exponent field. If you can arrange to call a Fortran routine from C, you might save yourself some trouble. For example, g77 on this code:

Code:

    program main

    d = 1.232
    x = -1.1500018e18

    write(6,9), d, x
9   format( "d = ", e14.6e3, " x = ", e20.10e4 )

    end

produces this:

Code:

% ./a.out
d =  0.123200E+001 x =  -0.1150001820E+0019

Best wishes ... cheers, makyo