I’ve implemented the simple 1d Gaussian example from intro to pyro thus:

```
def model(hparams, y=None):
prior_mean, prior_sd, obs_sd = hparams
theta = numpyro.sample("theta", dist.Normal(prior_mean, prior_sd))
return numpyro.sample("y", dist.Normal(theta, obs_sd), obs=y)
```

I can do posterior inference (of theta) in this using MCMC thus:

```
mu = 8.5; tau = 1.0; sigma = 0.75;
hparams = (mu, tau, sigma)
nuts_kernel = NUTS(model)
mcmc = MCMC(nuts_kernel, num_warmup=100, num_samples=1000)
mcmc.run(rng_key_, hparams, y)
```

Now I want to generate new samples from the unconditional joint. I tried this

```
data = model(hparams)
print(data)
```

but get this error:

What does this mean?

I also tried `model(rng_key, hparams)`

and `model(hparams, None)`

and `model(rng_key, hparams, None)`

but none of these work.