Syntax:
    NOISE: rerr={amount of white noise to add to predicted response}
Example:
    NOISE: rerr=0.1    !noise corresponds to 10% of the predicted response

Purpose: relative standard deviation associated with the model-predicted response.

Data Type: real 

Unit: {same as response variable}

Default: 0.1 (10% relative error; must use the -n command line option to activate)

Comments:  White noise is generated by Monte Carlo sampling the from a normal distribution with a specified variance.  The predicted response with noise is given by P_noise = P(1+r), where r is the sampled white noise.  The ISMLŽ DRNNOA routine is to generate pseudorandom numbers (r) from a standard normal distribution using an acceptance/rejection method.

Last updated: 10 June, 2011
Webmaster