compositional.mle1.0.2 package

BFGS Solver

Chain Solvers with Early Stopping

Clear derivative cache

Compose Multiple Function Transformations

Compose Multiple Solvers Sequentially

compositional.mle: Compositional Maximum Likelihood Estimation

Coordinate Ascent Solver

Backtracking line search

Golden section line search along one coordinate

Check if cli package is available

Create a progress handler for optimization

Finalize trace recorder into trace data

Fisher Scoring Solver

Get Fisher information function from problem

Get score function from problem

Gradient Ascent Solver

Sequential Solver Composition

Grid Search Solver

Check if solver converged

Check if object is an mle_constraint

Check if object is an mle_numerical

Check if object is an mle_problem

Check if tracing is enabled

L-BFGS-B Solver (Box Constrained)

Merge trace data from multiple results

Create domain constraint specification

Create an MLE Problem Specification

Create a Trace Configuration

Nelder-Mead Solver (Derivative-Free)

Create a trace recorder

Newton-Raphson Solver

Normal Sampler Factory

Get number of iterations

Extract Optimization Path as Data Frame

Elastic net penalty (combination of L1 and L2)

L1 penalty function (LASSO)

L2 penalty function (Ridge)

Plot Optimization Convergence

Plot Trace Data Directly

Print MLE Trace Data

Parallel Solver Racing (Operator)

Race Multiple Solvers

Random Search Solver

Record an iteration to trace

Simulated Annealing Solver

Uniform Sampler Factory

Conditional Refinement

Update an mle_problem

Verbose Output Utilities

Add penalty term to log-likelihood

Multiple Random Restarts

Create stochastic log-likelihood with subsampling