Closes #5316: SplitMix64RNG

[ajpotts]

Add stateless, index-based RNG via SplitMix64

Summary

This PR introduces a stateless, index-based random number generator to Arkouda using a SplitMix64 mixing function. The new implementation enables reproducible random sequences that are independent of execution order, parallelism, and locale count, making them suitable for sampling, shuffling, sketching, and other distributed algorithms.

The PR adds:

• A new Chapel module (SplitMix64RNG) implementing SplitMix64-based kernels
• Python-side generator APIs for stateless uint64 and uniform[0,1) generation
• In-place and allocating variants
• Comprehensive correctness and statistical test coverage

Motivation

Existing RNG paths in Arkouda rely on mutable state, which can make results sensitive to chunking and execution order. Many analytics and data-parallel algorithms instead need pure functions of (seed, stream, index).

Design & Implementation

• Chapel: new SplitMix64RNG module with fillRandU64 and fillUniform01 kernels
• Python: new internal Generator methods for stateless RNG
• Deterministic mapping: r[i] = splitmix64((start_idx + i) ^ key(seed, stream))

Testing

Adds extensive reproducibility, statistical, and independence tests, with conservative thresholds to avoid CI flakiness.

Compatibility

No changes to existing RNG behavior; new functionality is opt-in.

Future Work

Potential public API exposure and use in distributed shuffle/sampling paths.

Closes #5316: SplitMix64RNG

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
⚠️ Please upload report for BASE (main@9bedaa2). Learn more about missing BASE report.

Additional details and impacted files

@@           Coverage Diff            @@
##             main     #5323   +/-   ##
========================================
  Coverage        ?   100.00%           
========================================
  Files           ?         4           
  Lines           ?        63           
  Branches        ?         0           
========================================
  Hits            ?        63           
  Misses          ?         0           
  Partials        ?         0           

Flag	Coverage Δ
python-coverage	100.00% <ø> (?)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[1RyanK]

Looks good!

[1RyanK]

Is this supposed to be SplitMix64RNG.chpl?

[ajpotts]

Fixed.

[1RyanK]

Is there a reason these are uint64(int(...)) but in _fill_with_stateless_u64_from_index they are just uint64(...)?

[ajpotts]

Good catch. They should be the same. The reason was to ensure an int scalar was passed into uint64, but I think a better way to enforce that is to add @typechecked to the functions. I made the change.