Closed eddelbuettel closed 7 years ago
Expanded the second example.
A second examples shows the NaiveBayesClassifier
class.
#include <RcppMLPACK.h> // MLPACK, Rcpp and RcppArmadillo
#include <mlpack/methods/naive_bayes/naive_bayes_classifier.hpp> // particular algorithm used here
// [[Rcpp::depends(RcppMLPACK)]]
// [[Rcpp::export]]
arma::irowvec naiveBayesClassifier(const arma::mat& train,
const arma::mat& test,
const arma::irowvec& labels,
const int& classes) {
// MLPACK wants Row<size_t> which is an unsigned representation
// that R does not have
arma::Row<size_t> labelsur, resultsur;
// TODO: check that all values are non-negative
labelsur = arma::conv_to<arma::Row<size_t>>::from(labels);
// Initialize with the default arguments.
// TODO: support more arguments>
mlpack::naive_bayes::NaiveBayesClassifier<> nbc(train, labelsur, classes);
nbc.Classify(test, resultsur);
arma::irowvec results = arma::conv_to<arma::irowvec>::from(resultsur);
return results;
}
We need a quick helper function to get test data, again mimicking the unit tests:
#include <RcppMLPACK.h> // MLPACK, Rcpp and RcppArmadillo
#include <mlpack/methods/naive_bayes/naive_bayes_classifier.hpp> // particular algorithm used here
// [[Rcpp::depends(RcppMLPACK)]]
// [[Rcpp::export]]
Rcpp::List getData(const char* trainFilename, const char* testFilename) {
arma::mat trainData, testData;
mlpack::data::Load(trainFilename, trainData, true); // note implicit transpose
mlpack::data::Load(testFilename, testData, true);
// Get the labels, then remove them from data
arma::rowvec trainlabels = trainData.row(trainData.n_rows -1);
arma::rowvec testlabels = testData.row(testData.n_rows -1);
trainData.shed_row(trainData.n_rows - 1);
testData.shed_row(trainData.n_rows - 1);
return(Rcpp::List::create(Rcpp::Named("trainData") = Rcpp::wrap(trainData),
Rcpp::Named("testData") = Rcpp::wrap(testData),
Rcpp::Named("trainlabels") = trainlabels,
Rcpp::Named("testlabels") = testlabels));
}
Now that we can fetch the data from R, and use it to call the classifier:
rl <- getData("/home/edd/git/mlpack/src/mlpack/tests/data/trainSet.csv", # should add to RcppMLACK2
"/home/edd/git/mlpack/src/mlpack/tests/data/testSet.csv")
trainData <- rl[["trainData"]]
testData <- rl[["testData"]]
trainlabels <- rl[["trainlabels"]]
testlabels <- rl[["testlabels"]]
res <- naiveBayesClassifier(trainData, testData, trainlabels, 2)
## res was a rowvector but comes back as 1-row matrix
all.equal(res[1,], testlabels)
As we can see, the computed classification on the test set corresponds to the expected
classification in testlabels
.
matrix(c(1, 2, 3, 1, 2, 3), nrow=2, byrow=TRUE)
head(trainData)
...Was fighting with the data and found that whole aspect ... cumbersome. mlpack is a little weird as it tranposes.
I think next step is to bring that example data set into the package, with a help page etc pp. But not today.
Looks good. I'm happy to contribute more examples, if that would be helpful.
Personally, I'd like to see an example with the testing step done with a model that was previously trained and saved to disk. Like they do in the command-line tutorial: http://www.mlpack.org/docs/mlpack-2.1.1/doxygen.php?doc=lrtutorial.html#linreg_ex3_lrtut
Closing the ticket as the file was posted. Will file new issue on need for more docs etc pp
Poking @thirdwing @coatless @MHenderson : if you have a moment, can you read across this before I post this to the Rcpp Gallery. I have a two-day streak going on new posts, and the day off so I figured I may as well :) In all seriousness, I think it is time to beat the drum a bit more for RcppMLPACK{1,2} and to sort out where we go with RcppMLPACK2.
title: "RcppMLPACK2 and the MLPACK Machine Learning Library" author: "Dirk Eddelbuettel" license: GPL (>= 2) tags: machine_learning armadillo mlpack summary: "RcppMLPACK2 bring access to MLPACK to R"
mlpack
mlpack is, to quote, a scalable machine learning library, written in C++, that aims to provide fast, extensible implementations of cutting-edge machine learning algorithms. It has been written by Ryan Curtin and others, and is described in two papers in BigLearning (2011) and JMLR (2013). mlpack uses Armadillo as the underlying linear algebra library, which, thanks to RcppArmadillo, is already a rather well-known library in the R ecosystem.
RcppMLPACK1
Qiang Kou has created the RcppMLPACK package on CRAN for easy-to-use integration of mlpack with R. It integrates the mlpack sources, and is, as a CRAN package, widely available on all platforms.
However, this RcppMLPACK package is also based on a by-now dated version of mlpack. Quoting again: mlpack provides these algorithms as simple command-line programs and C++ classes which can then be integrated into larger-scale machine learning solutions. Version 2 of the mlpack sources switched to a slightly more encompassing build also requiring the Boost libraries 'program_options', 'unit_test_framework' and 'serialization'. Within the context of an R package, we could condition out the first two as R provides both the direct interface (hence no need to parse command-line options) and also the testing framework. However, it would be both difficult and potentially undesirable to condition out the serialization which allows mlpack to store and resume machine learning tasks.
We refer to this version now as RcppMLPACK1.
RcppMLPACK2
As of February 2017, the current version of mlpack is 2.1.1. As it requires external linking with (some) Boost libraries as well as with Armadillo, we have created a new package RcppMLPACK2 inside a new GitHub organization RcppMLPACK.
Linux
This package works fine on Linux provided mlpack, Armadillo and Boost are installed.
OS X / macOS
For maxOS / OS X, James Balamuta has tried to set up a homebrew recipe but there are some tricky interaction with the compiler suites used by both brew and R on macOS.
Windows
For Windows, one could do what Jeroen Ooms has done and build (external) libraries. Volunteers are encouraged to get in touch via the issue tickets at GitHub.
Example: Logistic Regression
To illustrate mlpack we show a first simple example also included in the package. As the rest of the Rcpp Gallery, these are "live" code examples.
We can then call this function with the same (trivial) data set as used in the first unit test for it:
Example: Naive Bayes Classifier
A second examples shows the
NaiveBayesClassifier
class.I also placed the (locally rendered) version here for now