iamgabeortiz / analytic_templates

A collection of folders and templates intended to provide structure and rapid prototyping for analytic projects using languages such as Python, R, SAS, SPSS, etc.

Project Description

A collection of folders and templates intended to provide structure and rapid prototyping for analytic projects using languages such as Python, R, SAS, SPSS, etc.

As of today (2014-05-11), this project is far from complete. R and SQL are my primary (and preferred) development platforms, thus most of the templates target these two languages. But, I would like to include templates for ANY language used for analysis. I will try to add a similar SAS project template by the end of the summer as I become more familiar with this widely used albeit restrictive language.

Organization

The Analytic Templates project is divided up into six parts.

• config
• file_templates
• project_templates
• snippets
• style_guides

Over time, each of these folders will become populated with language specific folders such as R, Python, SQL, etc.

config

Documents best practices for system configuration. This includes Emacs / Vim integration, Git, etc. Installation instructions and platform-specific gothas would also be appropriate.

file_tempaltes

Collection of language-specific file templates. Each template will target a specific task or analytic process and will assume you are using the corresponding project template. A file template is different from the project template in one very importany way. While the project templates will provide broad structure for an entire project, a file template will provide specific structure and boilerplate for a specific part of a project.

Example:

Most analysis projects result in some sort of output or report. Thus, there is a standard folder in the R project template called "reports/". However, I do NOT want to dictate or control what these reports should look like. Many of my projects result in reports in the form of Excel files or LaTeX files. Others may prefer to use Knitr, or they may just produce a CSV file. Any / all of these are reports and they are valid outputs of an analysis. Thus, there should be example templates for each of these methods to assist the user. All R projects should have a reports folder, but they should NOT require the user to enjoy using LaTeX.

project_templates

A collection of language-specific analytic project templates. The goal of a project template is to provide a common setup and configuration for analysis projects.

Example:

I always store data in a folder called data. I always store SQL queries in a folder called sql. In each of these folders I place a README.md file to document the files in the that folder. Reproducing this common boilerplate was time consuming and the inevitable small differences I introduced over time were annoying. From that frustration, this project was born.

snippets

Collection of useful, documented code snippets that demonstrate how to do a specific task. This part of the repo is nearly empty.

style_guides

Collection of language-specific style guides. I have not yet started this part of the project.