Attributes: Defining Your Data
Definition: A quality or characteristic describing an observation, such as colour or size, equivalent to feature in machine learning contexts.
Attributes in Code vs. AI
- In OOP: An attribute is a property of a class (
User.email). - In AI/ML: An attribute (or feature) is an input variable (
Age,Income). - In Vibe Coding: An attribute is a specific detail you want the AI to include.
Prompting for Attributes
When asking AI to generate data or mock objects, you must be explicit about the attributes.
- Bad Prompt: “Generate some users.”
- Result:
[{name: "John"}, {name: "Jane"}]
- Result:
- Good Prompt: “Generate 5 users with attributes:
UUID,hashed_password,last_login_iso_date, andsubscription_tier(enum: free, pro).”- Result: Perfect, production-ready JSON.
Feature Engineering with AI
If you are doing ML, you can ask the AI to “suggest attributes.”
- Scenario: You want to predict house prices.
- Prompt: “I have a dataset of houses. What derived attributes (features) should I create to improve my model?”
- AI Answer: “Create ‘price_per_sqft’, ‘age_of_roof’, and ‘distance_to_school’.”
The “Vibe” of Data
In the vibe coding era, you often deal with unstructured text. You can use AI to Extract Attributes.
- Prompt: “Read this email and extract the attributes:
sender_sentiment,urgency_level, andaction_items.”
This turns messy human text into clean structured data for your code to consume.