Attribute Error on asv bechmarking

[Lakshyadevelops]

Attribute Error on asv benchmarking. Iine 16 in module benchmarks/solarposition.py

To Reproduce
Steps to reproduce the behavior:

1. conda create new environment using ci/requirements-py3.7-min.yml
2. install pvlib,asv
3. cd pvlib-python/benchmarks
4. asv quickstart (recognizes existing template) (optional)
5. asv run

Expected behavior
benchmark should run

Screenshots

Versions:

• pvlib.__version__: 0.9.5
• pandas.__version__: 1.5.3
• python:" 3.9.13 but fails in asv condapyenv for all py 3.7-3.10

Additional context

Since I was using version 0.9.5>= 0.6.1 I could comment out the the function and asv benchmarks run without problem.

[kandersolar]

It shouldn't be trying to evaluate the second branch (the one with get) unless you're using a very old pvlib, or there is some issue with determining the version string. I suggest modifying that benchmark file to report the pvlib version it is using so you can inspect it.

FWIW when I am running benchmarks myself, I have gotten into the habit of using asv run --python=same so it uses my currently activated environment instead of the ones that asv manages. That makes it a little easier to know exactly what's going on.

conda create new environment using ci/requirements-py3.7-min.yml

FYI we do not use the environments in the ci folder for benchmarking purposes. It might work, but it's not tested or supported.

[Lakshyadevelops]

shouldn't the environment made by asv used for standardized benchmarking? ci environment is my working environment not benchmarking environment.

running source code with asv run does not succeed in building environment because of scipy version. I am opening a PR for it.

#1704

I also noticed you support 3.7-3.10 & 3.11 without numba but asv only used 3.7-3.8 for benchmarking.
Would it be interesting to benchmark from 3.9-3.11. Specially 3.11 with it's new improvements as numba release candidate is coming next week?
@kanderso-nrel
PS-any way to determine pvlib version in asv environment?