Initialzing NUTS with covariance matrix as mass matrix

@junpenglao when I execute this code I get the error:

TypeError
Traceback (most recent call last)
in
----> 1 basic_model.logp_dlogp_function([alpha, sigma]).dict_to_array()

TypeError: dict_to_array() missing 1 required positional argument: ‘point’

What is the ‘point’ argument I have to pass?