Design and Implementation of a Quant Database (v2)
Last week, we shared discussions in considerations for designing a quant database. This week, we take those considerations and implement a database solution in Python with MongoDB Time-Series collections.
We obtained a fully-functional query-write database that handles both time-series data and non-time-series data. The discussions here are both meaningful and functional. We also integrated the database service with the data retrieval service we talk about earlier.
This version of the paper adds 30 pages of discussion and code implementation of a functional database service. We intended to include some more code, but the discussions got fairly detailed - we would include more advanced methods in the next version.
A Link to Previous Posts:
A preview of NEW additions IN BLUE sections (TOC): (download full PDF report below)
In the next iteration, we want to consider optimizations to increase read and write throughput. We will cover topics in asynchronous programming, multi-threading, batch processing, reducing network trips and HTTP handshakes,
Full Report (Paid Readers):
Free readers of HangukQuant who wishes to get one-time access to the paper should reach out directly.
READ: if you spot any errors, please contact me directly and I will see to it. Note that the code implementation written is not taken from a production environment and then adapted - they are written `on the fly’ and managing thousands of lines in new code every few days is a non-trivial task. There are bound to be errors and we need you to be our eyes.
FAQ: Can you contribute to the project? If you have any suggestions, feel free to DM us.
FAQ: Can you send us the compilable code files? Yes, I am possibly doing that but only for paid readers at a later stage as part of a purchasable book, not at the paper level.