calcLikelihood: Calculate the likelihood for the samples in a MADproject...

In anchoredDistr: Post-Processing for the Method of Anchored Distributions

Description

calcLikelihood returns an updated MADproject with the likelihood values based on the observation and realization data in the MADproject or, optionally, a subset thereof.

Usage

calcLikelihood(proj, dsubset, num_realz = max(proj@realizations$rid),
  samples = 1:proj@numSamples)

## S4 method for signature 'MADproject,numeric'
calcLikelihood(proj, dsubset,
  num_realz = max(proj@realizations$rid), samples = 1:proj@numSamples)

## S4 method for signature 'MADproject,ANY'
calcLikelihood(proj, num_realz, samples)

Arguments

proj	The MADproject object with data read from the MAD# databases.
dsubset	The subset of inversion data to use for the likelihood calculations.
num_realz	The number of realizations to use in the likelihood calculation (defaults to all in the realizations slot)
samples	A vector of sample IDs for which to calculate likelihood values (defaults to all available in the realizations slot)

Details

The likelihood calculation utilizes the np package for non- parametric density estimation with all inversion data as dependent (i.e. multivariate likelihood distributions are estimated).

Value

proj The updated MADproject object with a filled likelihood slot.

Methods (by class)

• proj = MADproject,dsubset = numeric: Calculates the likelihood using a subset dsubset of inversion data zid

• proj = MADproject,dsubset = ANY: Calculates the likelihood using all inversion data zid

Examples

data(pumping)
pumping <- calcLikelihood(pumping, 100) #Inversion data as time step 100

anchoredDistr documentation built on May 2, 2019, 2:12 p.m.