JOSS Review Feedback
Hi @despot thank you for your hard work on this package! I will post feedback for your JOSS review on this thread going forward.
I have a few comments/questions I wanted to bring to your attention before I go deeper into the contents and functionality of pysewer:
-
I used pip to install pysewer and have a few notes: (a) in the pip installation instructions, please change
cd pysewer
tocd pysewer_dev
(b) I had a hard time getting all the packages (specifically gdal) installed using pip. I received the following error when I tried to use pip tp install pysewerERROR: Could not build wheels for gdal, which is required to install pyproject.toml-based projects
and had to use conda to install gdal and only then was able to use pip to install pysewer. It might be useful for other potential users who prefer using pip to be aware that they might run into problems installing gdal. (c) I am unable to run the example_sewer_network.ipynb file and run into this error: -
Since I cannot currently run the example ipynb file on my PC, I took a look at the version that has been stored on your repository. The output for cell [10] throws up several RuntimeWarnings: what causes those and do they prevent the user from being able to instantiating the model domain? If these warnings are benign, please consider adding text to this example file or to the docs explaining why and when a user would expect such an error. (Similarly, there seems to be a NameError in cell [12].)
-
Is it possible to customize the number of pumps in the network? E.g., can a user add a constraint that imposes a limit that only two or fewer pumps can be added?
-
The paper contains the sentence "Results are fully-connected sewer networks that connect all buildings to one or several predefined wastewater treatment plant (WWTP) locations." What does fully-connected mean here? I see in the jupyter notebook file that two separate and seemingly disconnected sewer networks are created.
-
As we briefly discussed on the main JOSS review thread, it would be great (in the open source spirit of the journal!) to provide a means for users not affiliated with UFZ to raise issues easily. Do you plan to mirror this repository on Github or provide a link to some sort of open forum?
Thanks! I will proceed with the rest of the review once I am able to run the example files on my machine.