/Kerblam_prototype

A Kerblam! toy project

Primary LanguageR

A Kerblam! toy project

This repository contains a minimal data analysis project as a personal reminder on how to use Kerblam!.

In this project, a single R script (getStats.R) can be launched from two different Bash pipelines (blue_pipe.sh and red_pipe.sh) for the analysis of the same data set (Roughness Adhesion Replicates.txt) made up of 42 observations of 12 variables each. The script will

  1. save the correlation matrix among the 12 variables as a TSV file,
  2. save the correlation pair-plot in PDF format,
  3. print on-screen the summary statistics resulting from a multiple linear regression analysis of the data.

The two pipelines only differ in the color used for the dots in the pair plot. However, unlike the blue_pipe, the red_pipe can be launched either locally or dockerized. In addition, both pipelines can be run on a reduced set of data (a subset of 4 variables) via the --profile test option.

Steps taken to create this Kerblam! project

  1. From my Coding directory (.)

    kerblam new Kerblam_prototype

  2. Set the following initial options

    Creating a new project in "Kerblam_prototype"!
Do you need Python? [yes/no]: n
Do you need R? [yes/no]: y
Do you want to use pre-commit? [yes/no]: n
Do you want to setup the remote origin of the project? [yes/no]: y
Enter your username: Feat-FeAR
What cloning method would you like? [ssh/https]: ssh
✅ "Kerblam_prototype" created!
✅ "Kerblam_prototype/data/in" created!
✅ "Kerblam_prototype/data/out" created!
✅ "Kerblam_prototype/src/pipes" created!
✅ "Kerblam_prototype/src/dockerfiles" created!
✅ "Kerblam_prototype/kerblam.toml" created!
✅ "Kerblam_prototype/.gitignore" created!
🏃 Executing 'git init'... Done!
🏃 Executing 'git remote add origin git@github.com:Feat-FeAR/Kerblam_prototype.git'... Done!

  3. From ./Kerblam_prototype

    • added data files to ./data/in
    • added the source code to ./scr (getStats.R)
    • added two alternative pipelines to ./scr/pipes (as Bash scripts)
    • added a dokerfile for one of the two pipelines to ./scr/dockerfiles
    • added a section [data.profiles.test] to kerblam.toml to run the pipelines on test data
    • added this README.md
    • tested everything
    • run kerblam data clean to remove output

  4. Pushed remotely on GitHub (creating the new remote repo at the same time)

    git add .
git commit -m "<message>"
gh repo create Kerblam_prototype --public --source=. --remote=upstream --push

Disclaimer

Important

This is in no way meant to be a guide to using Kerblam!, but just a small and innocent template for my future projects. Please refer to the official documentation for a guide on how to use Kerblam! and a full understanding of this same, albeit minimal, toy project.