particleFilter.estimates

Set or return the estimates for a particle filter

Syntax



obj = obj.estimates(Ye)


obj = obj.estimates(Ye, whichPrior)


[Ye, whichPrior] = obj.estimates


obj = obj.estimates(‘delete’)

Description



obj = obj.estimates(Ye)

Provides a set of observation estimates (Ye) for a particle filter. Overwrites any previously input estimates. Each row hold the estimates for a particular site, and each column holds the estimates for an ensemble member. Each element along the third dimension holds the estimates for a particular prior in an evolving set. If using a static prior, Ye should have a single element along the third dimension. Otherwise, Ye must have one element along the third dimension per assimilation time step - in this case, the particle filter will use the indicated set of estimates for each assimilated time step.



obj = obj.estimates(Ye, whichPrior)

Indicate which set of estimates to use in each assimilation time step. This syntax allows the number of sets of estimates (i.e. the number of priors) to differ from the number of time steps.



[Ye, whichPrior] = obj.estimates

Returns the current estimates for the particle filter object, and indicates which set of estimates (i.e. which prior) is used in each assimilation time step.



obj = obj.estimates(‘delete’)

Deletes any current estimates from the particle filter object.

Input Arguments

Ye

numeric array [nSite x nMembers x 1|*nTime*|*nPrior*]
Proxy estimates for the particle filter. A numeric array with one row per site, and one column per ensemble member. Each element along the third dimension holds the estimates for a unique prior. If using a static prior, Ye should have one element along the third dimension. If the number of elements along the third dimension matches the number of time steps, uses the appropriate set of estimates for each time step. If the number of priors is neither 1 nor the number of time steps, use the whichPrior input to indicate which set of estimates to use in each assimilation time step. NaN values are not permitted.

whichPrior

vector, positive integers [nTime]
Indicates which set of estimates to use in each assimilation time step. Must have one element per assimilation time step. Each element of whichPrior is the index of a particular prior - essentially, the index of an element along the third dimension of Ye.

Output Arguments

obj

scalar particleFilter object
The particleFilter object with updated estimates

Ye

numeric matrix [nSite x nMembers x 1|*nTime*|*nPrior*]
The current estimates forthe particleFilter object. If you have not provided estimates, an empty array.

whichPrior

vector, positive integers [nTime] | []
Indicates which set of estimates (i.e. which prior) is used in each assimilation time step. If there is a static prior, returns an empty array.