gnoise

gnoise (num [, RNG])

The gnoise function returns a random value drawn from a Gaussian distribution such that the standard deviation of an infinite number of such values would be num.

gnoise returns a complex result if num is complex or if it is used in a complex expression. See Use of gnoise With Complex Numbers below.

The random number generator is initialized using a seed derived from the system clock when Igor starts. This almost guarantees that you will never get the same sequence twice. If you want repeatable "random" numbers, use SetRandomSeed.

The Gaussian distribution is achieved using a Box-Muller transformation of uniform random numbers.

The optional parameter RNG selects one of three pseudo-random number generators used to create the uniformly-distributed random numbers providing the input to the Box-Muller transformation.

If you omit the RNG parameter, gnoise uses RNG number 3, named "Xoshiro256**". This random number generator was added in Igor Pro 9.00 and is recommended for all new code. In earlier versions of Igor, the default was 1 (Linear Congruential Generator).

The available random number generators are:

RNG	Description
1	Linear Congruential Generator by L'Ecuyer with added Bayes-Durham shuffle. The algorithm is described in Numerical Recipes edition 2 as the function ran2(). This RNG has nearly 2³² distinct values and the sequence of random numbers has a period in excess of 10¹⁸.
	This RNG is provided for backward compatibility only. New code should use RNG=3 (Xoshiro256**).
2	Mersenne Twister by Makoto Matsumoto and Takuji Nishimura. It is claimed to have better distribution properties and period of 2¹⁹⁹³⁷-1. See M. Matsumoto and T. Nishimura, "Mersenne Twister: A 623-dimensionally equidistributed uniform pseudorandom number generator", ACM Trans. on Modeling and Computer Simulation Vol. 8, No. 1, pp.3-30 1998. More information is available at
	http://en.wikipedia.org/wiki/Mersenne_twister.
	This RNG is provided for backward compatibility only. New code should use RNG=3 (Xoshiro256**).
3	Xoshiro256** by David Blackman and Sebastiano Vigna. It has a period of 2²⁵⁶-1, is 4 dimensionally equidistributed and passes all statistical tests at the time of writing. For technical details, see "Scrambled linear pseudorandom number generators", 2019 (http://vigna.di.unimi.it/ftp/papers/ScrambledLinear.pdf).

Use of gnoise With Complex Numbers

gnoise returns a complex result if num is complex or if it is used in a complex expression.

Function DemoGNoiseComplex()
	// enoise(rv) in a complex expression returns a complex result
	// and is equivalent to cmplx(gnoise(rv),gnoise(rv))
	Variable rv = 1	// A real variable
	SetRandomSeed 1
	Variable/C cv1 = gnoise(rv)	// Real parameter, complex result
	SetRandomSeed 1
	Variable/C cv2 = cmplx(gnoise(rv),gnoise(rv))	// Equivalent
	Print cv1, cv2
	
	// gnoise(cv) is equivalent to cmplx(gnoise(real(cv)),gnoise(imag(cv)))
	Variable/C cv = cmplx(1,1)	// A complex variable
	SetRandomSeed 1
	Variable/C cv3 = gnoise(cv)	// Complex parameter, complex result
	SetRandomSeed 1
	Variable/C cv4 = cmplx(gnoise(real(cv)),gnoise(imag(cv)))	// Equivalent
	Print cv3, cv4
End

gnoise (num [, RNG])​

Use of gnoise With Complex Numbers​

See Also​

gnoise (num [, RNG])

Use of gnoise With Complex Numbers

See Also