{"Package":"tableMatrix","Title":"Combines 'data.table' and 'matrix' Classes","Version":"0.82.0","Authors@R":"c(person(\"Petr\", \"Lenhard\", email = \"petr.lenhard@inferencetech.com\", role = c(\"aut\", \"cre\")),\nperson(\"Inference Technologies\", email = \"mail@inferencetech.com\", role = \"cph\"),\nperson(\"Petra\", \"Hudeckova\", email = \"hudeckovapetra@gmail.com\", role = \"ctb\"))","Description":"Provides two classes extending 'data.table' class. Simple\n'tableList' class wraps 'data.table' and any additional structures together.\nMore complex 'tableMatrix' class combines 'data.table' and\n'matrix'. See for more\ninformation and examples.","Depends":"R (>= 3.3.1)","License":"GPL-2","Encoding":"UTF-8","LazyData":"true","RoxygenNote":"6.0.1","Imports":"data.table","Suggests":"knitr, rmarkdown","VignetteBuilder":"knitr","NeedsCompilation":"no","Packaged":"2018-06-26 15:20:03 UTC; contumax","Author":"Petr Lenhard [aut, cre],\nInference Technologies [cph],\nPetra Hudeckova [ctb]","Maintainer":"Petr Lenhard ","Repository":"CRAN","Date/Publication":"2018-06-26 20:24:01 UTC","repoType":"cran","tarballUrl":"ftp://cran.r-project.org/pub/R/src/contrib/tableMatrix_0.82.0.tar.gz","jsonAuthors":[{"name":"Petr Lenhard","email":"petr.lenhard@inferencetech.com","maintainer":true},{"name":"Inference Technologies ","email":"mail@inferencetech.com","maintainer":false},{"name":"Petra Hudeckova","email":"hudeckovapetra@gmail.com","maintainer":false}],"readme":"\ntableMatrix package\n===================\n\ntableMatrix package provides two classes extending data.table class.\nSimple `tableList` class wraps data.table and any additional structures\ntogether. More complex `tableMatrix` class combines data.table and\nmatrix.\n\nInstallation\n------------\n\nFrom CRAN:\n\n``` r\ninstall.packages(\"tableMatrix\")\n```\n\nFrom github:\n\n``` r\n# install.packages(\"devtools\")\ndevtools::install_github(\"InferenceTechnologies/tableMatrix\")\n```\n\ntableList class\n---------------\n\n### Motivation\n\nGoal is to wrap data.table class and other structures together and\npreserve data.table behaviour.\n\n### Example\n\nCombine data and linear model into one object.\n\n``` r\ndata(chickwts)\n \n# Bundle chickwts data.frame together with a linear model\nTL <- tableList(chickwts, lm(weight~feed, chickwts))\n\n# tableList behaves like a data.table \nmean(TL[feed==\"casein\", weight])\n#> [1] 323.5833\n\n# Aid part of the tableList object carries the linear model\naid(TL)\n#> \n#> Call:\n#> lm(formula = weight ~ feed, data = chickwts)\n#> \n#> Coefficients:\n#> (Intercept) feedhorsebean feedlinseed feedmeatmeal feedsoybean \n#> 323.583 -163.383 -104.833 -46.674 -77.155 \n#> feedsunflower \n#> 5.333\n```\n\ntableMatrix class\n-----------------\n\n### Motivation\n\nLet’s have a dataset with the following structure: first set of columns\nof varying types is intented as meta data, second set of columns of the\nsame type is intended as main data. `tableMatrix` stores meta data as a\ndata.table and main data as a matrix. It also keeps track of dimensions\nof main data, thus allowing to combine rows of varying lengths into one\nobject. As in tableList, tableMatrix can carry any additional aid data.\n\n### Example\n\nWorking with bitmaps of different sizes. Datasets `images8By8` and\n`images10By10` contain 8x8 and 10x10 images in the form of vectors. For\neach row first three columns represent image meta data, remaining\ncolumns represent the image itself. For more information see\n`?images8By8`.\n\n``` r\n# Load datasets\ndata(images8By8)\ndata(images10By10)\n\n# Create a signle tableMatrix object from both datasets\n# First 3 columns used as meta data, the rest as main data with corresponding dimensions\nTM <- tableMatrix(list(images8By8, images10By10),\nlist(1:3, 1:3), list(c(4:ncol(images8By8)),c(4:ncol(images10By10))), list(c(8,8), c(10,10)))\n\n# Default print displays the table (meta data) part\nTM\n#> direction dimX dimY\n#> 1: down 8 8\n#> 2: down 8 8\n#> 3: down 8 8\n#> 4: down 8 8\n#> 5: down 8 8\n#> --- \n#> 176: both 10 10\n#> 177: both 10 10\n#> 178: both 10 10\n#> 179: both 10 10\n#> 180: both 10 10\n\n# Number of matrices stored in the matrix (main data) part\nlength(mat(TM))\n#> [1] 2\n\n# Dimensions of the matrix part\nmatDim(TM)\n#> tm.matN tm.matDim1 tm.matDim2\n#> 1: 1 8 8\n#> 2: 2 10 10\n\n# Aid part is empty\naid(TM)\n#> list()\n\n\n# Image data for first row\nimg <- getRow(TM, 1)\n\n# Restoring dimensions of the image\ndim(img) <- getRowDim(TM, 1)\n\n# Visualising the image\nimage(img, axes=F)\n```\n\n![](man/figures/README-unnamed-chunk-5-1.png)\n\nLet’s create a heat map\n\n``` r\n# Subsetting via bracket passed to the table (meta data) part\n# We choose first matrix type, down direction\nTM1down <- TM[.(1)][direction==\"down\"]\n\n# One matrix in the matrix part of TM1down\nlength(mat(TM1down))\n#> [1] 1\n\n# One dimension row\nmatDim(TM1down)\n#> tm.matN tm.matDim1 tm.matDim2\n#> 1: 1 8 8\n\n# Heatmap\nimgHeat <- colMeans(mat(TM1down, 1))\n\n# Restoring dimensions of the heatmap\ndim(imgHeat) <- getRowDim(TM1down, 1)\n\n# Visualising heatmap\nimage(imgHeat, axes=F)\n```\n\n![](man/figures/README-unnamed-chunk-6-1.png)","jobInfo":{"package":"tableMatrix","version":"0.82.0","parsingStatus":"success","parserVersion":1,"parsedAt":"2018-06-26T22:29:57+0000"}}

Description

Usage

Arguments

Value