rklees

Below the code: # Generate dataset measurement_error = st.MeasurementError(name="obs") IRW = st.LevelTrendComponent(order=2, innovations_order=[0, 1]) cycle = st.CycleComponent(name="annual_cycle", cycle_length=12, innovations=True) # cycle length is the period in number of samples; non-integer...

Samplers 'numpyro' and 'blackjax' run smoothly.

Software versions I use: python version = 3.11.2 (v3.11.2:878ead1ac1, Feb 7 2023, 10:02:41) [Clang 13.0.0 (clang-1300.0.29.30)] pymc version = 5.10.3 pandas version = 2.1.4 arviz version = 0.17.0 numpy version...

This does not solve the problem. Error messages persist.

I don't specify any coords. I set up a model as IRW = st.LevelTrendComponent(order=2, innovations_order=[0, 1]) annual_cycle = st.CycleComponent(name='annual_cycle', cycle_length=12, innovations=True) SA_cycle = st.CycleComponent(name='SA_cycle', cycle_length=5.347, innovations=True) obs_noise = st.MeasurementError(name="obs") pymc_mod...

Any update regarding this issue?

This problem still persists in version 0.1.1. Is there any hope that someone will fix it? Otherwise, I can't use pymc-experimental for unobserved components. Would be a pitty.

Here are the details: NotImplementedError Traceback (most recent call last) File [~/miniforge3/envs/BSSM/lib/python3.11/site-packages/pytensor/link/basic.py:102](http://localhost:8888/lab/tree/Software/Python%20software/PyMC%20experimental%20Notebooks/~/miniforge3/envs/BSSM/lib/python3.11/site-packages/pytensor/link/basic.py#line=101), in Container.__set__(self, value) 100 try: 101 # Use in-place filtering when[/if](http://localhost:8888/if) possible --> 102 self.storage[0] = self.type.filter_inplace( 103...

How to get a version with this issue fixed?

I will use the latest version of the example notebook and do some tests.