Help with something in C Programming.

[markbilly]

I have a file with a big long list of numbers in.

I want to count the number of lines in the file up to the number closest (rounded down) to one inputted by the user.

So say I have: "upper_lim" the number entered by the user.
                        "num_lines" the number of lines counted.
                        "num_on_line" the number on the file of the file.

I've done this while() loop, but I'm getting a nonsense number.

Code: ags

        while (num_on_line <= upper_lim) 
        {
                fscanf(input, "%f", &num_on_line);
		num_lines++;
	}

I've also tried this:

Code: ags

        while (!EOF) 
	{
               fscanf(input, "%f", &num_on_line);
		
               if ( num_on_line <= upper_lim )
               {
                      num_lines++;
               }
	}

Help! Why am I getting a nonsense number like 2.07e-310 ?

[Calin Leafshade]

Help! Why am I getting a nonsense number like 2.07e-310 ?

Thats not a nonsense number :p its standard form.

it mean 2.07 * 10^310.. i.e a very big number. your loop is getting stuck or something

EDIT: no wait!

2.07 * 10^-310 a very SMALL number.. oops

[markbilly]

I know what orders of magnitude are, I'm a Physicist! I meant in terms of the program it's nonsense. I doubt many data files have a number of lines < 1.

[Calin Leafshade]

lol oh.. :p

_{I'm a physicist too... a physicist of lurv}

..do you ever get called a physician?

[markbilly]

Yes. My brother (a social scientist) called me that once. He doesn't any more!

Any ideas on the program?

[bicilotti]

Not sure. Maybe:
fscan reads formatted things. Are you using fprintf to write the data?

Also, what does exactly give you the 2.07 * 10^-310 output? (A printf I suppose) Sure to have put %d and not %f in the first argument?
Is num_lines an int?

[Crimson Wizard]

As bicilotti sais, you maybe use wrong formatting option. %f is used for floating-point, for integer you should use %i or %d.
Yet, you did not explained where do you get this result exactly, so it is just a guess.

[markbilly]

num_lines is an int

upper_lim and num_on_line are floats

a printf("number of lines: %f \n", num_lines); gives the nonsense number.

[Calin Leafshade]

yeah your using the float formatting option for an int

%i

[markbilly]

The list of numbers are like this.

So they have to be scanned as floats, surely?

[Calin Leafshade]

well perhaps, but the number of lines is still an integer

you cant have half a line..

printf("number of lines: %d \n", num_lines);

[bicilotti]

a printf("number of lines: %f \n", num_lines); gives the nonsense number.

printf("number of lines: %d\n", num_lines); should work

Maybe you want to use cout << to get screen output, so to forget about % formatting  (though it's c++ lib., not c).

[markbilly]

It doesn't matter whether I use %i, %f, %g, whatever, it still returns a value of zero. Whether that is by displaying it as 2.7e-310 or 0.000000 or just 0.

There is something messing up with the loop and I can't work out what...

[Crimson Wizard]

Can you please show a full function here?
The problem can be anywhere, maybe it is not in the lines you already posted.

[Calin Leafshade]

i would guess your initial condition is coming out as false and thus the loop never gets run.

[Crimson Wizard]

For example, how EOF is initialized and where is it set after then?
Even if it is initialized properly, I don't see how it will change afterwards, thus from my point of view it may cause endless loop. 

[markbilly]

This is the full code, I'm get the value of the area under a function between 0 and the "upper_lim" i.e. upper limit of integration. The data file contains x values in one column and corresponding f(x) values in another column.

The code is rubbish. I'm not 100% sure what I'm doing.

Code: ags

#include <stdio.h>
#include <stdlib.h>
#include <math.h>

/*define function*/
float area_intg(float *fnc_fx, float *fnc_dx, int lines);

int main(int argc, char* argv[])
{   
	/*define a float for user-inputed upper limit of integration*/
    float user_input, upper_lim, answer;
	float *value_x, *value_xmin1, *dx, *fx, *fx_xmin1;
	
	/*Check the upper limit is enter via command-line correctly*/
    if (argc != 2)
    {
             printf("Need to enter upper limit of integration \n");
             exit(EXIT_FAILURE);
    }
    
	/*Assign the user's string input to the float "upper_lim"*/
    user_input = sscanf(argv[1], "%f", &upper_lim);
    
    /*upper_lim = atof(argv[1]);*/
    
    /*Check upper_lim is valid*/
    if (upper_lim == 0)
    {
     printf("Upper limit cannot be 0, this is the lower limit. \n");
     exit(EXIT_FAILURE);
    }
    if (upper_lim >= 1.57)
    {
     printf("Upper limit cannot exceed pi/2. \n");
     exit(EXIT_FAILURE);
    } 
    
	/*open the file containing data*/
	FILE *input;
	const char file_input[] = "px270prog3a.dat";
	input = fopen(file_input, "r");
	
	/*Find "num_lines" using while loop*/
	int num_lines = 0;
	float num_on_line;
	
	while (num_on_line <= upper_lim) 
	{
        fscanf(input, "%f", &num_on_line);
		num_lines++;
	}

    
	/*manually assign number of array arguments, using malloc*/
	value_x = (float*) malloc(sizeof(float) * num_lines);
	value_xmin1 = (float*) malloc(sizeof(float) * num_lines);
	fx = (float*) malloc(sizeof(float) * num_lines);
	dx = (float*) malloc(sizeof(float) * num_lines);
	
	/*scan file and store data in arrays*/
	int n = 0;
	
	for (n = 0; n < num_lines; n++)
	{
		fscanf(input, "%f %f", &value_x[n], &fx[n]);
		fscanf(input, "%f", &value_xmin1[n-1]);
		dx[n] = value_x[n] - value_xmin1[n-1];
	}
	
	answer = area_intg(fx, dx, num_lines);
	
	printf("number of lines: %i \nanswer to integral: %f \n", num_lines, answer);
}

float area_intg(float *fnc_fx, float *fnc_dx, int lines)
{
	float area;
	int j;

	for (j = 1; j < (lines); j++)
	{
           area += fnc_fx[j-1] * fnc_dx[j];
	}
	return(area);
}

[Crimson Wizard]

Two warnings:

1) num_on_line is used unitialized. Better write
float num_on_line =0.0; or something
2) in function area_intg variable area is used unitialized. Better write
float area = 0.0;

Otherwise you may get unexpected results. This depends on compiler you use; AFAIK Microsoft C compilers usually do not initialize variables with zeros.

By the way, since num_on_line is used in loop condition, it may make the error you have. So try to initialize it before loop and see what happen.

[markbilly]

That just returns area as zero all the time, now.

Getting output:

number of lines: 2
answer to integral: 0.000000

[Crimson Wizard]

Well, I ran you program with argument = 1.1 and got:

number of lines: 14007
nswer to integral: 677823296.000000