Mass matrix contains zeros on the diagonal and The derivative of RV is zero

Questions
1

Hi, Below is my model which I am trying to sample Posterior Predictive distribution.

with pm.Model() as model_g:

Ke = pm.Normal('Ke',0.25,0.1)

ϵ = pm.Normal('ϵ', 0,0.1)

# Parametrization for the (alpha) shape parameter.

Kej =  0.25 * np.exp(Ke)

X_pred = pm.Normal('X_pred', mu=onepk(X0,t2,Ke), sd=ϵ, observed=X_sim1)

trace_g = pm.sample(2000, tune=1000,cores=2)

My onepk function outputs X value as per,

def onepk(X0,time,Ke):

X = X0 *np.exp(-Ke * time)

return X

I am getting the below error with respect to derivative of my parameters.

image
image1718×872 128 KB

My observed data is between 1e-3 and 50, and the mean, std for my Ke parameter is obtained from previous simulation.

I know this is a regular issue but I am not able to fix this.

Any input is highly appreciated and Thank you.