Floating Point Calculations are not Standardized between Systems

[Grochocinski]

Describe the Bug

Simulations are not repeatable due to differences in handling floating point operations between operating systems.

Steps To Reproduce

1. Run the integrated test suite on a windows machine
2. Enter WSL
3. Run the exact same tests and some of them will now fail

Expected Behavior

Behavior should be standardized between different environments.

Environment

Please provide some details about your environment to help us replicate the bug:

• Operating System: Windows 11, Ubuntu 24.04
• Python Version: Python 3.11.10

Additional Details (If Applicable)

Float rounding can be non-deterministic and NumPy will switch between 32-bit and 64-bit floats by default depending on the system. The combination of these two factors can make reproducing data using identical configurations on two different systems impossible.

A good solution may be to standardize 32/64-bit for all values and to add some amount of deterministic rounding to eliminate non-deterministic properties from the results.

[Grochocinski]

Between the Testing refactor (#145) and the ASTRO-DF Refactor (#149), all tests should be passing on both Windows and Linux (at least with the test release of mrg32k3a 1.1). I'm sure that there will inevitably be some discrepancy in the future, but I'm going to close this issue since results on both platforms are identical for the time being...