Add bigframes.bigquery.rand() function

This change adds `bigframes.bigquery.rand()` which wraps the BigQuery `RAND()` function.
It accepts a Series or DataFrame as input to determine the shape and index of the output Series.
It includes a warning about non-determinism in the docstring.

Additionally, `SqlScalarOp` has been updated to accept an `is_deterministic` argument, which is set to `False` for `rand()`.

Tests are added in `tests/unit/bigquery/test_mathematical.py` and `tests/system/small/bigquery/test_mathematical.py`.

Ran `nox -s format lint mypy` and confirmed no issues were introduced.

Co-authored-by: tswast <247555+tswast@users.noreply.github.com>

Author: google-labs-jules[bot]

bigframes/bigquery/_operations/mathematical.py

@@ -28,8 +28,8 @@ def rand(input_data: Union[series.Series, dataframe.DataFrame]) -> series.Series
 
     .. warning::
         This method introduces non-determinism to the expression. Reading the
-        same column twice may result in different results. This value might
-        change and not to use this value or any value derived from it as a join
+        same column twice may result in different results. The value might
+        change. Do not use this value or any value derived from it as a join
         key.
 
     **Examples:**