:orphan: .. only available via index, not via toctree .. title:: Package Recipe 'r-graphstatsr' .. highlight: bash r-graphstatsr ============= .. conda:recipe:: r-graphstatsr :replaces_section_title: :noindex: A Shiny app to easily generate advanced graphics and perform some non\-parametric tests\, designed for metabolomics data analysis and developed by MetaBoHUB\-Metatoul. :homepage: https://forge.inrae.fr/etienne.rifa/graphstats :license: GPL-3.0-or-later :recipe: /`r-graphstatsr `_/`meta.yaml `_ .. conda:package:: r-graphstatsr |downloads_r-graphstatsr| |docker_r-graphstatsr| :versions: ``2.6.1-0`` :depends bioconductor-rhdf5: :depends r-base: ``>=4.4,<4.5.0a0`` :depends r-base64enc: :depends r-bit64: :depends r-car: :depends r-config: ``>=0.3.1`` :depends r-datamods: :depends r-dplyr: :depends r-dt: :depends r-factoextra: :depends r-ggplot2: :depends r-ggrepel: :depends r-ggstatsplot: :depends r-glue: :depends r-golem: ``>=0.3.1`` :depends r-gridextra: :depends r-htmltools: :depends r-openxlsx: :depends r-plotly: :depends r-pmcmrplus: :depends r-rcolorbrewer: :depends r-reshape2: :depends r-rlang: :depends r-shiny: ``>=1.6.0`` :depends r-shinyalert: :depends r-shinybs: :depends r-shinydashboard: :depends r-shinywidgets: :depends r-sortable: :depends r-stringr: :depends r-tibble: :depends r-tidyr: :depends r-waiter: :depends r-yaml: :requirements: :additional platforms: .. rubric:: Installation You need a conda-compatible package manager (currently either `micromamba `_, `mamba `_, or `conda `_) and the Bioconda channel already activated (see :ref:`set-up-channels`). While any of above package managers is fine, it is currently recommended to use either micromamba or mamba (see `here `_ for installation instructions). We will show all commands using mamba below, but the arguments are the same for the two others. Given that you already have a conda environment in which you want to have this package, install with:: mamba install r-graphstatsr and update with:: mamba update r-graphstatsr To create a new environment, run:: mamba create --name myenvname r-graphstatsr with ``myenvname`` being a reasonable name for the environment (see e.g. the `mamba docs `_ for details and further options). Alternatively, use the docker container:: docker pull quay.io/biocontainers/r-graphstatsr: (see `r-graphstatsr/tags`_ for valid values for ````) .. |downloads_r-graphstatsr| image:: https://img.shields.io/conda/dn/bioconda/r-graphstatsr.svg?style=flat :target: https://anaconda.org/bioconda/r-graphstatsr :alt: (downloads) .. |docker_r-graphstatsr| image:: https://quay.io/repository/biocontainers/r-graphstatsr/status :target: https://quay.io/repository/biocontainers/r-graphstatsr .. _`r-graphstatsr/tags`: https://quay.io/repository/biocontainers/r-graphstatsr?tab=tags .. raw:: html Download stats ----------------- .. raw:: html :file: ../../templates/package_dashboard.html Link to this page ----------------- Render an |install-with-bioconda| badge with the following MarkDown:: [![install with bioconda](https://img.shields.io/badge/install%20with-bioconda-brightgreen.svg?style=flat)](http://bioconda.github.io/recipes/r-graphstatsr/README.html) .. |install-with-bioconda| image:: https://img.shields.io/badge/install%20with-bioconda-brightgreen.svg?style=flat :target: http://bioconda.github.io/recipes/r-graphstatsr/README.html