Adrian Seyboldt
Adrian Seyboldt
## Description of your problem or feature request In numba mode, we can't compile functions that contain sigmoid operations: ```python import aesara.tensor as at from scipy import special import numba...
https://github.com/LLNL/sundials/blob/b78e575639babe99aea9a0558d0f64732d6d729e/src/cvodes/cvodea.c#L2037 Out of bounds access could easily happen here if a user accidentally changes the tolerance or user_data.
I am using the adjoint sensitivity analysis functionality of sunodes, and sporadically I get segmentation faults during the backward pass. Unfortunately I could not reproduce this with a small example...
I often use `plot_parallel` for finding clues about why divergences might be happening, there seem to be two problems with this however. First, it plots *the accepted point of a...
Currently the rhs function is defined using sympy and then compiled using numba. Now that aesara has a numba backend we can use aesara to compile this function instead, which...
`ChemicalProp.speciesStandardGibbsEnergies` seems to read invalid values in the first element of the array, each time I execute the following code the first value changes. ```python import reaktoro as rk db...
Metadata definitions that refer to themselves like `!0 = { !0 }` are allowed by llvm, and used to make some metadata entries globally unique (see for instance https://llvm.org/docs/LangRef.html#noalias-and-alias-scope-metadata). It...
I made the 0.10 release, but it might need a bit longer for this to appear on pypi. See https://github.com/PyO3/maturin/issues/1998
Right now everything in nutpie (the sampler and the logp functions) are working in float64. For some models we could get decent speedups if we also support float32. Pytensor can...
If a deterministic variable uses a RandomVariable, it won't properly update the seed in the `expanding_function` of the compiled pymc model, as we only use the numba function but do...