search

BeamIO-Inc / algorithm_toolkit

This project provides an easy way for researchers and developers to develop and share algorithms related to geospatial data and imagery.
https://algorithm-toolkit.readthedocs.io/en/latest/index.html
MIT License
6 stars 4 forks source link

Conda dev environment #36

Closed ryanlaclair closed 3 years ago

ryanlaclair commented 3 years ago

What?

A simple addition of an anaconda environment.yml file for use when doing development work on the algorithm toolkit.

Why?

BeamIO has seemed to standardize on conda for environment management across projects. Including an environment file in the repo ensures developers making changes to the algorithm toolkit can focus on the code rather than the setup.

Closes #35

ryanlaclair commented 3 years ago

Will this be a base for the example atk project as well? Or will the example project have its own yml file? If it is used as a base, we should maybe test to make sure Shapely works with Python 3.9. I actually did not even know Python 3.9 was released until now haha

Good point - I didn't originally intend to use it for the example project since that seems to use a regular requirements.txt file. However we could update the example project to use conda if we want. Even with an update to use conda in the example, we shouldn't need to use this as a base - the Algorithm Toolkit runtime dependency libraries should be automatically installed when the ATK is installed.

This environment file would be used for development work, where you cloning the code from github (does not isntall dependencies) rather than pip install (does install dependencies).

wrp5031 commented 3 years ago

Good point - I didn't originally intend to use it for the example project since that seems to use a regular requirements.txt file. However we could update the example project to use conda if we want. Even with an update to use conda in the example, we shouldn't need to use this as a base - the Algorithm Toolkit runtime dependency libraries should be automatically installed when the ATK is installed.

I think we should have a conda env yml file for the example project, but that is for another discussion.

chriswilley commented 3 years ago

Good point - I didn't originally intend to use it for the example project since that seems to use a regular requirements.txt file. However we could update the example project to use conda if we want. Even with an update to use conda in the example, we shouldn't need to use this as a base - the Algorithm Toolkit runtime dependency libraries should be automatically installed when the ATK is installed.

I think we should have a conda env yml file for the example project, but that is for another discussion.

I would support that, but we should keep requirements.txt in there also for people who don't use conda.