Colin Carroll comments

Results 81 comments of


                                            Colin Carroll

TensorFlow v2.14.0 breaks TensorFlow Probability at import

Thanks for pinging on this, @matthewfeickert ! Note that using `tfp-nightly` in the meantime should work.

TurncatedNormal gives wrong results sometimes

Hey! Thanks for opening this issue -- it looks like the problem is with the boundaries here, as we might expect ``` import numpy.testing as npt import scipy.stats as st...

Warnings, etc. about Tutorial: Probabilistic Regression

Thanks for the report! The warnings are caused by code in the notebook -- I can update those. I checked the quantitative differences too without finding any answer -- my...

distributions.mvn_diag using deprecated scale_identity_multiplier kwarg

thanks! i'm taking a look at this.

VI - Variational Inference support

Maybe a weird suggestion here, but my current thinking is that a good "standard data structure" for VI would be an object that supports, say, `.sample` and `.log_prob` (and `.log_prob_parts`,...

VI - Variational Inference support

Thanks, @sethaxen -- well put! Maybe two issues I'm still stuck on are 1. I guess it seems weird to me to throw away answers to some (often) analytic questions...

Integration with Inference Gym

Note you should be able to use ``` from inference_gym import using_jax as gym ``` via [this test](https://github.com/tensorflow/probability/blob/main/spinoffs/inference_gym/inference_gym/backends/jax_integration_test.py#L17C1-L18C1)

Integration with Inference Gym

> So for example, it would be nice for this to serve as a place to see how to run each algorithm with the best possible tuning For what it...

WIP: Add polynomial regression example

A bit of a side note that I thought this was really nice. I found the bit on orthogonal polynomials really interesting, perhaps as a separate notebook? In particular, I...

WIP: Add polynomial regression example

Also, my only productive feedback is that ``` df_falling.assign(tsquared=t**2) ``` is a neat trick to create a copy of `df_falling` with an extra column. I had trouble spotting it, but...