GCD approximation for type double numbers

[teleport]

Hi, I am doing a phys experiment, and I find myself trying to obtain some pattern of quantization of some measurements, i.e., I'm trying to find a number (double) that divides at least a significant portion of my data, with an arbitrary remainder. Does anyone know of any algorithm that does this for type double numbers? I really need your help on this one. Thanks a lot.

 

[CRGreathouse]

double mod(double n, double m) {
	if (n < 0)
		n = -n;
	if (m < 0)
		m = -m;
	int tmp = (int)(n / m);

	return n - tmp * m;
}

 

[ravenprp]

Hello,

Thank you for reaching out for help with your experiment. Finding the greatest common divisor (GCD) for type double numbers can be a bit challenging, as these numbers can have a large range and precision. However, there are a few approaches you can take to approximate the GCD for your data.

One option is to use the Euclidean algorithm, which is commonly used for finding the GCD of two integers. This algorithm can also be adapted for type double numbers by converting them to integers with a common scale. This approach may not give you an exact GCD, but it can provide a good approximation.

Another option is to use the continued fraction algorithm, which can also be adapted for type double numbers. This algorithm involves finding the continued fraction representation of your numbers and then using a recurrence relation to approximate the GCD. This approach may give you a more accurate approximation compared to the Euclidean algorithm.

I would also recommend looking into numerical libraries or packages that may have functions specifically designed for finding the GCD of type double numbers. These may provide more efficient and accurate solutions for your experiment.

I hope this helps and wish you the best of luck with your experiment. Let me know if you have any further questions or need any clarification.