Investigate

This is a command-line tool that initializes a directory for a scientific project. The directory has a standardized structure recommended for scientific projects (see below), and is tracked by git.

Usage:

$ # Make new Python project called "rna_expression"
$ investigate rna_expression --language python

The directory may optionally be initialized with a programming language (currently either "julia" or "python"):

• If "julia", add a Project.toml for your project, and an empty src/main.jl
• If "python", initialize a new conda env, and add environment.yml and an empty src/main.py

Directory structure

• raw: Raw data, e.g. experimental data, or data from external research groups. Should not be modified at all.
• src: This directory contains code and scripts used to reproduce the results. The file main.py or main.jl should produce all results using only the data in directories raw and choices.
• tmp: Directory for throwaway analyses and intermediate results. Anything in this directory should be able to be deleted with no big loss
• cache: Also for intermediate results, but for content that is troublesome to recreate, e.g. results of long-running simuations or long-running computation
• choices: For files that are not raw files, but impossible to recreate automatically, because they rely on humans (you!) making judgement calls.
• results: For final analytic results. main.jl/py should write results to this directory, primarily
• paper: For results related to submission of any papers, e.g. manuscripts or publication-ready figures.