by Manuel Ramon (mailto:[email protected]) on date Jan 4, 2017.
This is a draft directory structure to use when starting a new project. It is based on my own personal preferences and represents only a general structure which can be modified to fit the needs of each particular project.
All my research project contain the following five main directories:
bin: contains all the executablesdata: a place to put data files. There is no rules to organize this directory, although it is advisable to have arawsubdirectory in which we will put the original (raw) data files. It is also advisable to save original files as read-only to prevent data loss and name them in a way that is easy to identify what information the data files have.docs: this directory contains the bibliography resources to be used in the project, the papers resulted from the research activity, reports, talks, etc.results: results obtained from the different analyses, such as data outputs, tables, figures, etc.src: contains the source code of all programs, shell scripts, make files, etc.
When data files are shared between different projects, it is advisable to put them in a common data folder (outside the project directory) and create symbolic links to those data files in the data folder.
During the development of the project, many data files will be obtained by
the analysis of raw data. These data files can be stored within the data folder
if they are going to be use in subsequent analyses, or within the results folder
if they are final outputs.
Although not included in this template, it can be useful to include a test folder
in which save all the different temporary tests carried out during the development
of the project.