Skip to content

project-workflow.md

Professional Project Workflow Glossary

A reference for professional development practices.

Workflow

A defined sequence of steps for completing a task consistently and reproducibly.

Reproducibility

The ability to run the same project with the same inputs and obtain the same outputs. A core professional and scientific standard. Achieved through version control, dependency management, and deterministic code.

Deterministic Output

Output that is identical every time the program runs with the same inputs. Required for reproducible analytics. Avoid randomness unless explicitly seeded.

Run

Execute the program so it performs its full set of tasks: read inputs, compute results, write artifacts, log progress.

Verify

Confirm the project works as expected by running it and checking outputs and logs.

Debug

The process of finding and fixing errors in code. Use log messages, print statements, and error messages to identify the source of problems.

Logging

Writing timestamped messages during execution to record what the program did. Supports debugging, verification, and reproducibility. Messages are written to a log file and typically also to the console.

Log Level

A category indicating the importance of a log message. Standard levels in order of severity:

  • DEBUG: detailed diagnostic information
  • INFO: confirmation that things are working as expected
  • WARNING: something unexpected happened but execution continues
  • ERROR: a serious problem; part of the program failed
  • CRITICAL: a severe error; the program cannot continue

Log File

The file where log messages are written during a run. Typically named project.log and stored in the project root.

Evidence

Files that demonstrate the program ran correctly. Includes: output CSVs, log files, plots, screenshots.

Artifact

A file generated by the program as a result of a run. Often stored in artifacts/ or output/. Never manually edited. Artifacts and outputs are evidence.

Professional Workflow

The practice of working the way professionals do: using version control, managing dependencies, writing documented and reproducible code, and producing verifiable outputs.

Iteration

Improving a project through repeated cycles of run / evaluate / refine. Professional systems are rarely finished; they are often continuously improved.

Code Review

Structured evaluation of code by another person using defined criteria. Identifies bugs, unclear logic, missing documentation, and improvement opportunities.

Maintenance

The ongoing work required to keep a system accurate, functional, and trustworthy over time. Includes updating dependencies, adjusting thresholds, and refining logic as conditions change.