Planning before programming

Often, people start programming before having a general idea of what they want to accomplish. If you're an experienced programmer, this may be a good way to get a feel for the problem, since you have already developed intuition, and you'll probably end up throwing away the first couple of attempts anyway. However, if you're a novice programmer, I recommend you make your objectives clear and explicit before writing any code (putting them into writing can help). It will help you make better decisions by asking yourself how a certain way of doing things will affect your objectives. So, before we set up anything, we need to understand and make our general objectives explicit:

1. Understand the big picture of the analysis quickly.
2. Reproduce our analysis automatically by executing a single file.
3. Save all the resulting objects, text, and images for the analysis.
4. Measure the amount of time it takes to perform the full analysis.
5. When working on iterative processes, know the completed percentage.
6. Be able to find and change each part of the analysis easily.

To fulfill these general objectives, we need to develop modular code with well-managed dependencies that are flexible (easy to change) and friendly to side-effects (saving objects, texts, and images). Even if your explicit objectives don't require it, you should make a habit of programming this way, even when just doing data analysis.