PVInspect

This package provides methods for the analysis of solar modules using different imaging modalities. We put huge efforts into providing a clean and easy to use API and additional tooling like the wrapper classes for module images that bundle image with meta data in an expressive way. You are invited to publish you own methods using this API and hence contribute to building a consistent and open tooling that might be useful to others. If you like to integrate your methods directly, please create a merge request.

Installation

We recommend to install pvinspect in a Python virtual environment using Python 3.8. This document only supplies a short description on the usage of virtual environments in Python. For a complete guide, please refer to this tutorial.

0. Check that you have the correct Python version available:

python --version

Depending on the OS and configuration of your machine, python needs to be substituted with python3, python3.exe or even the absolute path to the Python interpreter. In any case, the displayed version should be compatible to pvinspect. At the moment, pvinspect requires Python 3.10.

1. Create a new virtual environment:

python -m venv /path/to/your/venv

This initializes a new virtual environement in /path/to/your/venv (which you should adapt to your needs). All packages that you install within this environment are stored in this folder. Depending on the OS and configuration of your machine, python needs to be substituted with python3 or python3.exe.

2. Activate the new virtual environment:

# For Windows users (Powershell):
/path/to/your/venv/Scripts/Activate.ps1

# For Linux users:
source /path/to/your/vent/Scripts/activate

Please note that the procedure depends on the type of shell that you are using. We've given example for Powershell and Linux-shell users.

3. Install pvinspect:

pip install git+https://github.com/ma0ho/pvinspect.git@master

This installs the pvinspect package in the currently active environment (which is normally indicated in the shell). So please make sure to have the correct environment activated. After installation, you can use pvinspect at any time from within that environment.

A note to Anaconda users

We do not ship pvinspect as an Anaconda package. However, you can of course install it in an Anaconda environment using pip. However, you might come across an error indicating that geos_c.dll is missing. In that case, you need to install shapely using conda:

conda install -c conda-forge shapely

Usage

This package contains example notebooks that demonstate the most common use cases.

Issues

In case you encounter anything that does not work, please open an issue and provide a precise description (include your OS version, python distribution and the like) as well as a minimal code example to reproduce the issue.

Known issues

In case you install pvinspect in a conda environment using pip on Windows, the shapely library, which is installed as a dependency of pvinspect does not find geos.dll. This can be fixed by installing shapely using conda:

conda install shapely

This is reported as a bug to shapely: #1032

Acknowledgement

We greatly acknowledge the HI-ERN for providing us a demo dataset of electroluminescense images that is published along with this package.

We gratefully acknowledge funding of the Federal Ministry for Economic Affairs and Energy (BMWi: Grant No. 0324286, iPV4.0).

ma0ho / pvinspect