Sequential Simulation, Multiple-Variable Case

1. template
   <
     class GeovalueIterator,
     class ForwardIterator,
     class Cdf,
     class CdfEstimator,
     class MarginalCdf,
   >
   inline int
   sequential_cosimulation(
     GeovalueIterator begin, GeovalueIterator end,
     ForwardIterator first_neigh, ForwardIterator last_neigh, 
     Cdf& ccdf,
     CdfEstimator& estim,
     MarginalCdf& marginal
   )
    
   
   
2. template
   <
     class GeovalueIterator,
     class ForwardIterator,
     class Cdf,
     class CdfEstimator,
     class MarginalCdf,
     class Sampler
   >
   inline int
   sequential_cosimulation(
     GeovalueIterator begin, GeovalueIterator end,
     ForwardIterator first_neigh, ForwardIterator last_neigh, 
     Cdf& ccdf,
     CdfEstimator& estim,
     MarginalCdf& marginal,
     Sampler& samp
   ) 
   
   
   

This function performs a sequential simulation of the range of geo-values delimited by iterators begin and end, accounting for multiple variables. At each location u being simulated, the conditional cdf is estimated based on the primary information stored in the first neighborhood *first_neigh[0], and the secondary information contained in the other nb_of_neighborhoods-1 neighborhoods (first_neigh is an array of pointers to neighborhoods). If at a given location no neighboring data is found, a new value is drawn from the marginal cumulative distribution: marginal.

In version 1, a new value is simulated using Monte-Carlo simulation: a probability is determined randomly and used to draw a realization from conditional cdf ccdf.

Version 2 allows to modify the way a simulated value is drawn from the ccdf.

The value returned is the number of problems that occured during the simulation.

Where defined

In header file <simulation.h>

Preconditions

• The range [begin,end) is a valid range.

• first_neigh[n] is a ``pointer'' to the (n+1)-th neighborhood to be accounted for. (n < nb_of_neighborhoods).

• estim and ccdf do not conflict: if estim is designed to estimate Gaussian cdf's, ccdf should be a Gaussian cdf.

Requirements on types

• GeovalueIterator is a model of Geo-Value Iterator.
• Neighborhood is a model of Neighborhood.
• Cdf is a model of CDF.
• MarginalCdf is a model of CDF. It can be different from cdf.
• CdfEstimator is a model of Cdf Estimator.
• Sampler (in version 2) is a model of Sampler.