Dopamine optical sensor. When dopamine is bound, it floreses and can detect micromolar changes and dopamine concentration.

See also Software Development Methodologies

documentation

• Comments
• Readme
• Wiki

specification

• UX
• UI
• High-Level Architecture (libraries, external APIs)
• Low-Level Architecture (modules, functions, internal APIs)

commenting

• Almost anything hardcoded (constants, strings, etc.)
• Anything confusing, tricky, nonstandard
• Historical notes: if something is added/removed, write it down
• TODO for bugs or hacks

README Files

• Best used as a quick-start guide
• What are key pieces of info they will pieces of info they will need?
  • What is your code supposed to do?
  • How does someone run your code?
  • How does a new engineer get set up?
• General overview of how things are laid out, with links to wiki pages with details

Wiki

• In-depth explanation of subsystems and modules
• Separate pages for each subsystem
• Include decisions of their design decisions
• Discussions of why systems are not designed differently

UI/UX Spec

• How do we know what the software is supposed to do?
• Varying levels of resolution
  • User stories
  • All the way up to granular details of UI elements

Don’t forgot to document defaults!

The dopamine circuitry in NF1.

Genetically encoded “sensors” to measure circuits.

There is extreme noise in the patient annotations: for instance, in their health rhythm labels there’s around 20% contradictions in the dataset.

1. accurate diagnosis
2. limiting “domain rule violations”

approach

1. take your dataset, and validate rules
2. IF validation is successful, train model normally with that sample
3. IF the validation is unsuccessful, then use the output samples as negative examples