To help Eric resolve the "Object reference not set to an instance of an object" error, the following steps and information may be useful:

1. Debugging Tools:

   • QuantConnect provides various debugging tools, including a built-in debugger, which allows stepping through code line-by-line to inspect variables and understand the program's internal state.
   • Logs and debug statements can be used to capture critical decision points and errors during the algorithm's execution.
   • Consider using self.log() or self.debug() to add messages in your code for tracking execution flow and state changes.

2. Understanding Coding Errors:

   • There are different types of coding errors: build errors, runtime errors, and logic errors.
   • The error message mentioned is related to a runtime error, often due to accessing an object that hasn't been initialized properly.

3. Error Message and Stack Trace:

   • The specific error "Object reference not set to an instance of an object" suggests a possible attempt to use a None object in an unexpected way. Without explicit user code details in the stack trace, it can be helpful to systematically check areas where objects are created and used (like adding securities or orders).

4. Code Validation:

   • The initial code snippet shared doesn't exhibit any standard errors that may lead to the error mentioned. However, the actual strategy might have more complexity which isn't visible here.
   • Consider using a minimal version of the algorithm for testing to isolate the issue.

5. PEP 8 Code Correction:

   • Ensuring the code adheres to Python’s PEP 8 for consistency and readability might help mitigate errors by making code more understandable.

By applying detailed logging, using the debugger, and examining the potential areas where objects might be improperly referenced or initialized, Eric should be able to pinpoint and resolve the runtime error.

Sources: Debugging - QuantConnect.com, Debugging Tools - QuantConnect.com

I've never seen any debug log messages actually showing up in QC. How do I see these when the backtest fails?

Note that the code runs fine for several periods, and fails later, during the backtest, so it's not easy to step through to find the error; it takes a long time to reach the fail point.