[zachisi]

Hi,

I have basic programming knowledge and have so far made a good start to my first programming assignment but have now hit a snag. I can honestly say I have no idea on where to go with this and my class notes and research have been of little aid (just can't get the jist of it). By the way im writing code in C language. Im a physics nerd deep down and programming just does not swing my way. Any help in helping me understand would be greatly appreciated (links, source material, step by step analysis/working would help a lot). Cheers guys

Q) After declaring the array, generate a smooth transition/gradient between two colours across the 256 wide by 128 high image. Build the colour values into your code to start with, but bear in mind that the finished version will require the user to supply them as 2 sets of 3 values (red, green and blue) between 0 and 255. Range checking should be implemented to ensure that colours outside this range are rejected.

So far, the code ive got to is:

#include <stdio.h>
int main()
{
	int pixelarray[128][256];
	int x, y;
	FILE *image;                                               /* File pointer */
	
	for (x = 0; x <= 128; x++) {                              /* Nested for loops to contain zero elements */
		for (y = 0; y <= 256; y++);}         

	image = fopen("image.ppm", "w");                            /* Creates file to be written */
	fprintf(image, "P3\n");                                     /* Writes to the created file */
	fprintf(image, "# image.ppm\n");
	fprintf(image, "128 256\n");
	fprintf(image, "255\n");
	fprintf(image, "%i", pixelarray[x][y]);                    /* Writes the array into the file */
	fclose(image);                                             /* Closes the file */
	return 0;
}

Thanks

[OLH064]

You shouldn't use a lesser-than-or-equal-to (<=) on line 8 and 9, but just lesser-than (<)
By using <=, you access the 128^th and 256^th elements of arrays that end at the 127^th and 255^th elements.

You'll have to make a loop like a fixed 8&9 to write 16 correctly.

Which direction should the gradient be in?
\ this way?
or | that way?

[ishkabible]

well whats your issue? you have to tell us or we can't help you. also post any errors your getting.

in the mean time...

gradients are simply change in color with respect to a location. for instance; circular gradients changes the color with respect to distance form a point. liner gradients change the color with respect to the x or y access(or both if the gradient goes across). you could plug in almost any mathematical function you could think of(even something like sin or cos).

you may want to look at this

[zachisi]

OLH064, on 26 January 2012 - 04:41 PM, said:

You shouldn't use a lesser-than-or-equal-to (<=) on line 8 and 9, but just lesser-than (<)
By using <=, you access the 128^th and 256^th elements of arrays that end at the 127^th and 255^th elements.

You'll have to make a loop like a fixed 8&9 to write 16 correctly.

Which direction should the gradient be in?
\ this way?
or | that way?

Thanks for that, ive changed it now. The gradient should be | this way.

[OLH064]

He wants a linear gradient, and some help understanding how to write this program.
...Or this is a "gimme te codez" instance.
You should work on getting the numbers, and then the gradient.
How should you go about getting the input?
How should you loop to get the gradient?
How many things should there be in the loop?

Spoiler

int R1, G1, B1;
int R2, G2, B2;
printf("gimme RGB twice in decimal format, separated by new lines!\n"); /* gotta say somefin' */
scanf("%i", &R1); /* says to scanf, I need a decimal from text entry; goes to R1 "What?" at address 0x-------- "oh." */

/* YOUR LOOP HERE */
array[x][y].R = (int)(255.0d * LinearInterpolate(R1, R2, ((double)y)/255.0d)); /* must be between 0.0 and 1.0, translated to double and back to int. MUST BE DOUBLE, NOT FLOAT. */
/* STOP LOOPING, FOOL */

Do you see what's missing?

double LinearInterpolate(
   double y1,double y2,
   double mu)
{
   return(y1*(1-mu)+y2*mu);
}

from Interpolation methods
Written by Paul Bourke
December 1999

This post has been edited by OLH064: 26 January 2012 - 05:10 PM

[zachisi]

ishkabible, on 26 January 2012 - 04:44 PM, said:

well whats your issue? you have to tell us or we can't help you. also post any errors your getting.

in the mean time...

gradients are simply change in color with respect to a location. for instance; circular gradients changes the color with respect to distance form a point. liner gradients change the color with respect to the x or y access(or both if the gradient goes across). you could plug in almost any mathematical function you could think of(even something like sin or cos).

you may want to look at this

The issue lies in i don't know where to go from now as i am struggling to grasp what to do. Thank you for the info

[ishkabible]

Then you need to identify what it is your struggling with

[zachisi]

ishkabible, on 26 January 2012 - 05:15 PM, said:

Then you need to identify what it is your struggling with

Im struggling with creating the code which will create a linear gradient through the array that ive created. Any help would be great

[OLH064]

Well, in your loop, you'll have to go though R, G, and B separately.

You'll have to use some funky math to get the value between the two Rs, Gs, or Bs that's influenced by the Y loop/array placement.

This post has been edited by OLH064: 26 January 2012 - 05:26 PM

[ishkabible]

Quote

Im struggling with creating the code which will create a linear gradient through the array that ive created. Any help would be great

What have you tried? Is what you tried not working? Programing is about breaking things down into parts; which part are you having trouble with? are you having trouble devising an algorithm or implementing it(hint: that's the first step)?

This post has been edited by ishkabible: 26 January 2012 - 06:29 PM

[zachisi]

ishkabible, on 26 January 2012 - 06:25 PM, said:

Quote

Im struggling with creating the code which will create a linear gradient through the array that ive created. Any help would be great

What have you tried? Is what you tried not working? Programing is about breaking things down into parts; which part are you having trouble with? are you having trouble devising an algorithm or implementing it(hint: that's the first step)?

Im having trouble with the actual algorithm for this, any sort of help would be great

[OLH064]

http://paulbourke.ne.../interpolation/
Has this neat linear interpolation function, but it interpolates between two doubles, and it takes the location as a double.

Ishkabible, could you translate

double LinearInterpolate(double y1,double y2,double mu)
{return(y1*(1-mu)+y2*mu);}

to integers?

[ishkabible]

ya, there are no divisions so it would work just fine.

better yet you could use templates...

template<class T>
T LinearInterpolate(T y1, T y2, T mu) {
   return y1 * (1 - mu) + y2 * mu;
}

now it works for any number type! gotta love me some generic code

[OLH064]

Yeah, mu is supposed to be between 0 and 1, and the 1-mu thing is to invert... it...

int linear(int y1, int y2, int mu, int murange)
{return (y1 * (murange - mu) + y2 * mu)/murange;} /* mu is between 0 and murange */

Somebody backup this post.

EDIT: y2 was named y1 when being passed to the function

This post has been edited by OLH064: 26 January 2012 - 08:41 PM

[ishkabible]

Quote

between 0 and 1

Oops, didn't think about that.

yes, yours appears to work the way I would expect(except you have 2 arguments named 'y1'. I renamed the second to y2). However, you do have to make sure that y1 and y2 are both divisible by murange however. if you don't you get a loss of precision.

effectively, mu and murange just act as a ratio to represent mu

This post has been edited by ishkabible: 26 January 2012 - 08:28 PM