Bayesian posterior sampling proble,

He guys when i run this code for posterior sampling
Covid19 = pm.Model()

with pm.Model() as Covid19:
tt.config.exception_verbosity=‘high’
theano.compile.mode.Mode(linker=‘py’, optimizer=None)
lambda_t_log = sociald_sigmoids(change_points, len(smoothed) + 16, pr_median_lambda=0.4)
new_I_t = SPEQIR(lambda_t_log, len(smoothed) +16, smoothed, 4900000)
cases = delay_cases(new_I_t, len(smoothed)+16, len(smoothed), 16)
sigma_obs = pm.HalfCauchy(‘sigmaobs’, beta=30, shape = len(smoothed))
a = tt.abs_(cases[:len(smoothed)]+ 1) ** 0.5* sigma_obs
pm.StudentT(name=‘name_student_t’, nu=4, mu=cases[:len(smoothed)], sigma= a, observed =smoothed)
Wehn i run this in i get this error for NUTS sampling
ValueError: expected an ndarray

During handling of the above exception, another exception occurred: