Python Streaming Data Cleaning | Dataford Interview Questions

Write a Python function to process and clean a streaming dataset before it hits our inference endpoint at Berkeley Research Group.

Implement clean_stream(records), where records is an iterable of dictionaries with id, text, and features fields. Return a list containing valid records with trimmed, lowercased IDs, trimmed text, and unchanged finite numeric features. Skip malformed records, empty values, invalid features, and duplicate IDs after normalization, keeping the first occurrence.

Constraints

1 <= len(records) <= 9999
Each valid record contains an ID string, a text string, and a non-empty list of numeric features
IDs are normalized by trimming whitespace and converting to lowercase
Duplicate normalized IDs keep their first valid record
Feature values must be finite integers or floating-point numbers

Function Signature

def clean_stream(records):

Your question is Python Streaming Data Cleaning. Start with the requirements in the Question tab.

Run and submit as often as you like. When you're ready, talk me through your approach or go straight to the code.

Problem

Write a Python function to process and clean a streaming dataset before it hits our inference endpoint at Berkeley Research Group.

Constraints

1 <= len(records) <= 9999
Each valid record contains an ID string, a text string, and a non-empty list of numeric features
IDs are normalized by trimming whitespace and converting to lowercase
Duplicate normalized IDs keep their first valid record
Feature values must be finite integers or floating-point numbers

Function Signature

def clean_stream(records):

Your question is Python Streaming Data Cleaning. Start with the requirements in the Question tab.

Run and submit as often as you like. When you're ready, talk me through your approach or go straight to the code.