Skip to content

Latest commit

 

History

History
102 lines (78 loc) · 3.64 KB

README.md

File metadata and controls

102 lines (78 loc) · 3.64 KB
Raw

Companion repository for ECAS-SFdS 2023 school

Welcome to the ECAS-SFdS 2023 school Random forests: basics, extensions and applications.

In this repository, you will find all the documents associated to the school and notably the different files needed for practicals sessions. There is one folder per speaker.

NEWS:

  • Nathalie’s material for the school is ready. Check instructions in nathalie/README.md!
  • Robin's material for the school is ready. Check description in robin/README.md!

INSTRUCTIONS:

Here is some tasks that we recommend you to perform before the school (the main reason is that the internet speed available there could be quite moderate):

  1. Clone this github repository on your computer by using the green button <> Code (for installation/use of git you can follow this Happy Git with R). You can also download the ZIP archive of the repository (but in this case you will have to re-download it in case of an update).

  2. Download the courses materials as described in the different folders.

  3. Download the practicals materials and/or data as described in the different folders.

  4. Install R (https://cloud.r-project.org/) and the following R packages (we recommend the use of the RStudio IDE):

    • ggplot2
    • grf
    • igraph
    • mlbench
    • PRROC
    • randomForest
    • randomForestSRC
    • ranger
    • reshape2
    • rfPermute
    • SISIR
    • GENIE3 (BioConductor)
    • RLT GitHub Version 4.2.5

    you can use the following commands (from within R) to do that:

    install.packages(c("ggplot2", "grf", "igraph", "mlbench", "PRROC", "randomForest", "randomForestSRC",
"ranger", "reshape2",  "rfPermute", "SISIR", "BiocManager", "remotes"))
BiocManager::install("GENIE3")                   
remotes::install_github("teazrq/RLT")

    for further information and in case of issues of installation of the RLT package, follow Install the RLT package.

    On linux, some system dependencies (C++ librairies) might be needed too.

  5. Install Python (we recommend Miniconda or Anaconda) and the following Python libraries:

    • notebook
    • jupyterlab
    • matplotlib
    • numpy
    • pandas
    • pyts
    • session_info
    • scikit-learn
    • rfpimp
    • xgboost
    • shap

    you can use the following command (from a terminal on Linux/MacOS or from 'Anaconda prompt', accessible from the Start menu, on Windows) to do that:

    pip install notebook jupyterlab matplotlib numpy pandas pyts session_info scikit-learn rfpimp xgboost shap

Alternatively, you can use cloud solutions, which prevent you from installing R or Python on your computer and can also help you resolve some installation issues:

  • Posit cloud for which you need to create a Posit account (please note that the free account might not be sufficient for all practical sessions)

  • Google Colab for which you need a Google account

You may still have to install the different packages/libraries once you’re logged in (depending of the type of document/project shared by the speakers). We recommend that you do it before the class has started.