Part 5/11:
An example API call involved sending a JSON payload that defined:
The collection name (e.g., "test")
The vector metric (initially dot product)
The vector size (e.g., 4 dimensions)
Sample commands demonstrated the nuances of JSON formatting differences between Windows and Linux, emphasizing the importance of escaping characters correctly.
Adding Data Points and Performing Searches
Once the collection was created, the next critical operation was adding data points (also called vectors). This involved sending JSON payloads containing:
Unique IDs for each data point, or allowing the engine to auto-generate IDs
The vector data itself
Optional payloads with arbitrary metadata