{"Package":"ivaBSS","Type":"Package","Title":"Tools for Independent Vector Analysis","Version":"1.0.0","Date":"2022-05-03","Authors@R":"c(person(given = \"Mika\", family = \"Sipil\", role = c(\"aut\", \"cre\"), email = \"mika.e.sipila@student.jyu.fi\"),\nperson(given = \"Klaus\", family = \"Nordhausen\", role = c(\"aut\"),\ncomment = c(ORCID = \"0000-0002-3758-8501\")),\nperson(given = \"Sara\", family = \"Taskinen\", role = \"aut\",\ncomment = c(ORCID = \"0000-0001-9470-7258\")))","Imports":"stats, graphics, BSSprep","Suggests":"LaplacesDemon","Encoding":"UTF-8","Maintainer":"Mika Sipil ","Description":"Independent vector analysis (IVA) is a blind source separation (BSS) model where several datasets are jointly unmixed. This package provides several methods for the unmixing together with some performance measures. For details, see Anderson et al. (2011) and Lee et al. (2007) .","License":"GPL (>= 3)","RoxygenNote":"7.1.2","NeedsCompilation":"no","Packaged":"2022-05-19 08:16:26 UTC; developer","Author":"Mika Sipil [aut, cre],\nKlaus Nordhausen [aut] (),\nSara Taskinen [aut] ()","Repository":"CRAN","Date/Publication":"2022-05-19 17:00:02 UTC","repoType":"cran","tarballUrl":"ftp://cran.r-project.org/pub/R/src/contrib/ivaBSS_1.0.0.tar.gz","jsonAuthors":[{"name":"Mika Sipil","email":"mika.e.sipila@student.jyu.fi","maintainer":true},{"name":"Klaus Nordhausen","email":null,"maintainer":false},{"name":"Sara Taskinen","email":null,"maintainer":false}],"readme":"## R Package for Independent Vector Analysis\n\nIndependent vector analysis (IVA) is a blind source separation (BSS) model where several datasets are jointly unmixed. This package provides several methods for the unmixing together with some performance measures.\n\n## How to install the package?\n\nMake sure you have git installed and clone the package using:\n\n`git clone https://github.com/mikasip/IVA.git`\n\nor just download `ivaBSS_1.0.0.tar.gz` file from this repository.\n\nMake sure you have R in your environment variables, open command prompt and run:\n\n`R CMD INSTALL path_to_file/ivaBSS_1.0.0.tar.gz`\n\n## How to use?\n\nThe package is used to estimate source vectors by unmixing the observed mixtures. The next example generates mixtures from sources following multivariate Laplace distribution and unmixes them using Newton update based IVA with multivariate Gaussian source density model.\n\n```\nif (require(\"LaplacesDemon\")) {\n # Generate sources from multivariate Laplace distribution\n P <- 4; N <- 1000; D <- 4;\n S <- array(NA, c(P, N, D))\n \n for (i in 1:P) {\n U <- array(rnorm(D * D), c(D, D))\n Sigma <- crossprod(U)\n S[i, , ] <- rmvl(N, rep(0, D), Sigma)\n }\n \n # Generate mixing matrices from standard normal distribution\n A <- array(rnorm(P * P * D), c(P, P, D))\n \n # Generate mixtures\n X <- array(NaN, c(P, N, D))\n for (d in 1:D) {\n X[, , d] <- A[, , d] \\%*\\% S[, , d]\n }\n \n # Estimate sources and unmixing matrices\n res_G <- NewtonIVA(X, source_density = \"gaussian\")\n }\n}\n\n```","jobInfo":{"package":"ivaBSS","version":"1.0.0","parsingStatus":"success","parserVersion":1,"parsedAt":"2022-05-19T19:29:53+0000"}}

Description

Arguments

Details

References