{"Package":"fnets","Type":"Package","Title":"Factor-Adjusted Network Estimation and Forecasting for\nHigh-Dimensional Time Series","Version":"0.1.6","Authors@R":"c(\nperson(\"Matteo\", \"Barigozzi\", email = \"matteo.barigozzi@unibo.it\", role = \"aut\"),\nperson(\"Haeran\", \"Cho\", email = \"haeran.cho@bristol.ac.uk\", role = c(\"cre\", \"aut\")),\nperson(\"Dom\", \"Owens\", email = \"dom.owens@bristol.ac.uk\", role = \"aut\"))","Maintainer":"Haeran Cho ","Description":"Implements methods for network estimation and forecasting of high-dimensional time series\nexhibiting strong serial and cross-sectional correlations under a factor-adjusted vector autoregressive model.\nSee Barigozzi, Cho and Owens (2024+) for further descriptions of FNETS methodology and\nOwens, Cho and Barigozzi (2024+) accompanying the R package.","Depends":"R (>= 4.1.0)","Imports":"lpSolve, parallel, doParallel, foreach, MASS, fields, igraph,\nRColorBrewer","License":"GPL (>= 3)","Encoding":"UTF-8","LazyData":"true","RoxygenNote":"7.3.1","Suggests":"testthat (>= 3.0.0)","Config/testthat/edition":"3","NeedsCompilation":"no","Packaged":"2024-01-23 08:17:08 UTC; h","Author":"Matteo Barigozzi [aut],\nHaeran Cho [cre, aut],\nDom Owens [aut]","Repository":"CRAN","Date/Publication":"2024-01-23 09:30:03 UTC","repoType":"cran","tarballUrl":"ftp://cran.r-project.org/pub/R/src/contrib/fnets_0.1.6.tar.gz","jsonAuthors":[{"name":"Matteo Barigozzi","email":"matteo.barigozzi@unibo.it","maintainer":false},{"name":"Haeran Cho","email":"haeran.cho@bristol.ac.uk","maintainer":true},{"name":"Dom Owens","email":"dom.owens@bristol.ac.uk","maintainer":false}],"readme":"# fnets\n\nContains methods for network estimation and forecasting for high-dimensional time series under a factor-adjusted VAR model. See \n\n> _fnets_: An R Package for Network Estimation and Forecasting via Factor-Adjusted VAR Modelling\n\nby Dom Owens, Haeran Cho and Matteo Barigozzi [arXiv:2301.11675](https://arxiv.org/abs/2301.11675) accompanying the R package, and\n\n> _FNETS_: Factor-adjusted network estimation and forecasting for high-dimensional time series\n\nby Matteo Barigozzi, Haeran Cho and Dom Owens [arXiv:2201.06110](https://arxiv.org/abs/2201.06110) for details of the methodology.\n\n\n## Installation\n\nTo install `fnets` from CRAN:\n\n```\ninstall.packages(\"fnets\")\n```\n\n\nTo install from GitHub:\n\n```\ndevtools::install_github(\"https://github.com/Dom-Owens-UoB/fnets\")\n```\n\n## Usage\n\nWe can generate an example dataset used in the above paper for simulation studies, by separately generating the factor-driven common component and the idiosyncratic VAR process as\n```\nset.seed(123)\nn <- 500\np <- 50\ncommon <- sim.unrestricted(n, p)\nidio <- sim.var(n, p)\nx <- common$data + idio$data\n```\n\nFit a factor-adjusted VAR model with `q = 2` factors and `lasso` for VAR transition matrix estimation\n```\nout <- fnets(x, q = 2, var.order = 1, var.method = \"lasso\", do.lrpc = FALSE)\n```\n\nPlot the Granger network induced by the estimated VAR transition matrices:\n```\nplot(out, type = \"granger\", display = \"network\")\n```\n\nEstimate and plot the partial-correlation and long-run partial correlation-based networks:\n```\nplrpc <- par.lrpc(out)\nout$lrpc <- plrpc\nout$lrpc.method <- 'par'\nplot(out, type = \"lrpc\", display = \"heatmap\")\n```\n\nEstimate the (long-run) partial correlation-based networks directly using `fnets`:\n```\nout <- fnets(x, q = 2, var.order = 1, var.method = \"lasso\", do.lrpc = TRUE)\n```\n\nForecast `n.ahead` steps:\n```\npr <- predict(out, n.ahead = 1, common.method = \"restricted\")\npr$forecast\n```\n\n\n\n\n\n","jobInfo":{"package":"fnets","version":"0.1.6","parsingStatus":"success","parserVersion":2,"parsedAt":"2024-01-23T11:30:13+0000"}}

Description

Usage

Arguments