Files vs Folders

We use a consistent structure to ensure reproducibility. In Data Science, being “organized” isn’t just about being neat—it is a requirement for your code to work on someone else’s machine.

The Golden Rule of Health Data

We treat data like a Crime Scene—never touch the evidence.

Start: data/ (The Crime Scene). READ ONLY.
- Rule: Never open a dataset in Excel and “fix” a typo. That is tampering with evidence.
- Correct Way: Load the “dirty” data into R/Python and fix the typo using code. This leaves a trail of exactly what you changed.
End: output/ (The Evidence). All graphs and cleaned tables go here.
The Detective: Your code (.qmd) moves information from Start to End.

Files vs. Folders (The Mental Model)

Files (e.g., report.qmd): These are sheets of paper. You write on them.
Folders (e.g., data/): These are labeled boxes. You store things in them.

Create a folder

In Explorer:

Right-click anywhere in the blank space of the Explorer pane.
Select New Folder
Name it:

images

Pro Tip: Always use lowercase and no spaces for folder names (e.g., my_data, not My Data). Computers struggle with spaces in file paths.

Typical structure

Your project should always look like this:

data/      raw data  
output/    generated results  
images/    screenshots  
.qmd       reports

Remember: The data folder is for input. The output folder is for results. They should never mix.