Learn R Programming

README

CRF is an R package implemented modeling and computational tools for conditional random fields (CRF) model as well as other probabilistic undirected graphical models of discrete data with pairwise and unary potentials.

How to install?

  • Stable version: The stable version of CRF is available in the following websites:

  • Latest beta version: The latest developmental version of CRF can be downloaded from GitHub and installed from source.

How to use?

There are documentation available within package. To read the documents in the R terminal using the command ?CRF

Some examples can be found in the 'tests' directory of package source.

R Packages that uses CRF

If you use CRF in your package and project, please inform us so that we can add the link here.

Copy Link

Version

Install

install.packages('CRF')

Monthly Downloads

84

Version

0.4-3

License

GPL (>= 2)

Issues

Pull Requests

Stars

Forks

Maintainer

LingYun Wu

Last Published

December 1st, 2019

Functions in CRF (0.4-3)

Chain

Chain CRF example
clamp.reset

Reset clamped CRF
Loop

Loop CRF example
clamp.crf

Make clamped CRF
CRF-package

CRF - Conditional Random Fields
Clique

Clique CRF example
Rain

Rain data
Tree

Tree CRF example
crf.nll

Calculate CRF negative log likelihood
crf.update

Update CRF potentials
decode.block

Decoding method using block iterated conditional modes algorithm
decode.tree

Decoding method for tree- and forest-structured graphs
decode.icm

Decoding method using iterated conditional modes algorithm
decode.greedy

Decoding method using greedy algorithm
decode.trbp

Decoding method using tree-reweighted belief propagation
decode.lbp

Decoding method using loopy belief propagation
decode.chain

Decoding method for chain-structured graphs
decode.conditional

Conditional decoding method
decode.marginal

Decoding method using inference
infer.tree

Inference method for tree- and forest-structured graphs
decode.rbp

Decoding method using residual belief propagation
decode.sample

Decoding method using sampling
Small

Small CRF example
decode.ilp

Decoding method using integer linear programming
decode.junction

Decoding method for low-treewidth graphs
train.mrf

Train MRF model
infer.cutset

Inference method for graphs with a small cutset
infer.conditional

Conditional inference method
make.par

Make CRF parameters
make.features

Make CRF features
duplicate.crf

Duplicate CRF
decode.cutset

Decoding method for graphs with a small cutset
decode.exact

Decoding method for small graphs
infer.exact

Inference method for small graphs
infer.junction

Inference method for low-treewidth graphs
sample.gibbs

Sampling method using single-site Gibbs sampler
make.crf

Make CRF
sample.exact

Sampling method for small graphs
sample.tree

Sampling method for tree- and forest-structured graphs
sample.junction

Sampling method for low-treewidth graphs
infer.chain

Inference method for chain-structured graphs
infer.sample

Inference method using sampling
get.potential

Calculate the potential of CRF
mrf.update

Update MRF potentials
get.logPotential

Calculate the log-potential of CRF
sample.chain

Sampling method for chain-structured graphs
infer.rbp

Inference method using residual belief propagation
mrf.stat

Calculate MRF sufficient statistics
mrf.nll

Calculate MRF negative log-likelihood
infer.lbp

Inference method using loopy belief propagation
infer.trbp

Inference method using tree-reweighted belief propagation
sample.cutset

Sampling method for graphs with a small cutset
train.crf

Train CRF model
sub.crf

Make sub CRF
sample.conditional

Conditional sampling method