BUG: MultivariateNormalQMC with specific QMCEngine and d == 1 erroneously throws exception

Describe your issue.

With the way the conditional is written in the lines below https://github.com/scipy/scipy/blob/838cfbe51a2aecaba19b24813769ef5aeebfe056/scipy/stats/_qmc.py#L1548 when initializing an MultivariateNormalQMC instance with d == 1 and an engine that is unquestionably an instance of scipy.stats.qmc.QMCEngine, we get the ValueError below. Works fine for d > 1.

Reproducing Code Example

from scipy.stats.qmc import MultivariateNormalQMC, Sobol

input_dim = 1
engine = Sobol(d=input_dim)
sampler = MultivariateNormalQMC(mean=np.zeros(input_dim), engine=engine)

Error message

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-5-f2da58e7ebf7> in <module>
----> 1 sampler = MultivariateNormalQMC(mean=np.zeros(1), engine=engine)

~/.virtualenvs/foo/lib/python3.8/site-packages/scipy/stats/_qmc.py in __init__(self, mean, cov, cov_root, inv_transform, engine, seed)
   1248             self.engine = engine
   1249         else:
-> 1250             raise ValueError("`engine` must be an instance of "
   1251                              "`scipy.stats.qmc.QMCEngine` or `None`.")
   1252 

ValueError: `engine` must be an instance of `scipy.stats.qmc.QMCEngine` or `None`.

SciPy/NumPy/Python version information

1.7.1 1.19.5 sys.version_info(major=3, minor=8, micro=10, releaselevel=‘final’, serial=0)

[ltiao]

Thanks @tupui for taking a look at this. I understand your previous point and think there is probably a case to be made for both approaches. Ultimately, I’d be happy with either option. Unfortunately, I am not participating in the PyData Sprint and also won’t have much time these days to contribute. Thanks for maintaining the project and keep up the good work!

[reshamas]

current:

       if engine is None:
            self.engine = Sobol(d=engine_dim, scramble=True, seed=seed)  # type: QMCEngine
        elif isinstance(engine, QMCEngine) and engine.d != 1:
            if engine.d != d:
                raise ValueError("Dimension of `engine` must be consistent"
                                 " with dimensions of mean and covariance.")
            self.engine = engine
        else:
            raise ValueError("`engine` must be an instance of "
                             "`scipy.stats.qmc.QMCEngine` or `None`.")

proposed:

       if engine is None:
            self.engine = Sobol(d=engine_dim, scramble=True, seed=seed)  # type: QMCEngine
    
        elif isinstance(engine, QMCEngine):
            if engine.d != d:
                raise ValueError("Dimension of `engine` must be consistent"
                                 " with dimensions of mean and covariance.")
            self.engine = engine
            if engine.d == 0:
              raise ValueError("xxx  0 not valid")
            if engine.d == 1:
              raise ValueError("xxx  1 not valid")
        else:
            raise ValueError("`engine` must be an instance of "
                             "`scipy.stats.qmc.QMCEngine` or `None`.")
         

if d=0, d=1, are there tests?