Giordon Stark
Giordon Stark
You're hitting this limitation in jax.
The following code words for me now: ```python # gotta patch stuff import sys from unittest.mock import patch # let's get started import pyhf jax_backend = pyhf.tensor.jax_backend(precision='64b') pyhf.set_backend(jax_backend) import jax.numpy...
For the jax issue that i temporarily patched, I filed the report here google/jax#3919
Are we saving a lot of extra code with the addition of another library here? It almost looks 1-1. It's not clear to me if we benefit strongly from this...
I still want to see something like this happen, to be honest.
> I gave this a try and this feature is extremely useful. It helps a lot with handling the mapping between `model.config.par_order` and the actual "flattened" parameters expected by `pyhf.pdf.Model.expected_data`...
JIRA ticket for cvmfs: [SPI-2221](https://sft.its.cern.ch/jira/browse/SPI-2221)
Tutorial update: pyhf/pyhf-tutorial#29
This is #1978
#993 seems to do the same thing.