Smit Lunagariya

This issue is fixed by https://github.com/QuantEcon/QuantEcon.py/pull/573, and @mmcky is working on the build system in #608. Probably, after that, we won't require `setup.py`.

> Not sure why this `job` is being skipped I guess because of this: https://github.com/QuantEcon/QuantEcon.py/blob/97cebb4c5ceb720b801c8587f463ea6df2660992/.github/workflows/conda_ci.yml#L69

How about changing that line to: `if: github.event.ref_type == 'tag' `

The reason for the incorrect result is this line: https://github.com/sympy/sympy/blob/2573a46a40c20dcd5cd05cc114343f82a2f3869f/sympy/stats/drv_types.py#L114-L115 This can be fixed in the following way: ```diff diff --git a/sympy/stats/drv_types.py b/sympy/stats/drv_types.py index c0fb01946f..786f7d13ec 100644 --- a/sympy/stats/drv_types.py +++ b/sympy/stats/drv_types.py...

Yes, sure. There is an open PR for Discrete RVs. If you wish you could start working on Continuous RVs

I agree that `pdf` should be renamed as `pmf`.

> the `density` shouldn't be allowed for a PMF either. `density()` is the only way the users could know the pmf/pdf of the distribution. It is like a common umbrella...

> to dispatch on different types of the `x` argument in `deepcopyop_2` when we were previously creating distinct functions for each argument type. Yes, exactly. `numba`'s jit should be able...

We need to discuss more on this approach and see how the functions are pickled in `dill`. The test failure on CI is very different from what it is on...

Hmm, this gets tricker when we start digging at a low level. IMHO, we can do the following: 1. Use the previous approach of moving static functions out and for...