Skip to content

PVGIS 6 vs 5

This section compares in detail PVGIS version 6 with its predecessor version 5. The transition to a new generation of PVGIS entails a complete software overhaul, including design, architecture, and functionality, while aiming to preserve all capabilities of the previous generation.

Target audience

Scope & Target audience

Category 5 6
Citizens Yes Yes
Prosumers Yes Yes
Researchers Indirectly & partly available through r.sun (GRASS GIS) Yes
Developers - Yes
Industry - -

Licensing

License 5 6
Use license european-union.europa.eu/legal-notice_en EUPL-1.2, CCSA (Documentation)

Architecture

Implementation Details

Component 5 6
Language C/C++ Python
Core Software Monolithic design Modular, clear API / CLI / Web API boundaries, scalable architecture
CLI Few independent standalone programs Unified CLI with rich options
API - Comprehensive API suite
Web API Flask-based FastAPI, asynchronous support
CORS - Yes / Work-in-progress
Aspect 5 6
Coding Style Mixed, inconsistent, extensive duplication PEP-8 compliant, Formmatted via Black
Testing Manual tests only Automated with continuous integration

Functionality

Capabilities & Functionality Comparison

Functionality 5 6
Solar Position Internal, Only the solar altitude part of final analysis, Hofierka 2002 (Jenco 1992) Selectable: Jenco (1992), NOAA, NREL SPA3, Skyfield, pysolar
Solar Incidence Internal, Hofierka 2002 (Jenco 1992) Selectable: Jenco (1992), Iqbal 1992
Solar Irradiance Part of final analysis, Hofierka 2002 Independent for each radiation component, Hofierka 2002
Photovoltaic Performance Huld 2011, Gracia 2016 Huld 2011, Gracia 2016

Scientific Functionality

Photovoltaic system 5 6
Grid-connected Yes Yes
Off-Grid Yes Planned
Tracking Yes Planned

Other capabilities

Capabilities 5 6
Solar irradiance time series Yes Yes
Meteorological time series Yes Yes
TMY Yes Work-in-progress
Fingerprinted results - Yes
QR-Code shareable output - Yes

Data Input/Output

File Formats Supported 5 6
Input formats PVGIS native time series format, Text Extensive support including any Xarray-supported format*2 like NetCDF & Zarr, CSV, JSON
Output formats Text, mixed/partly separated fields Rich output formats possible including CSV, JSON, XML, NetCDF, PNG
Database Databaseless1 Databaseless1

Performance

Efficienct & Reliability

Metric 5 6
Speed Instant Instant, Optimizable via Numba, DaCe
Reliability Known issues, occasionally affecting output Improved with robust testing
Accuracy Satisfactory checked with research work High, with rigorous testing
Automated Testing None Comprehensive suite via Pytest
Code Quality Not standardized PEP-8 compliant, linted code
Modularity Mixed code styles Fully modular with clean separation of concerns
Flexibility Limited, extensive hardcoding Highly flexible and customizable
Web API Flask-based Rebuilt with FastAPI for asynchronous operations
Core API - Robust core API supporting various integrations
CLI Few command line tools, basic UX Advanced CLI with Typer, rich UX
Data Handling PVGIS-native time series format only Supports multiple formats via Xarray
Plotting Results Via Web GUI only Integrated plotting in CLI and Web API

Customisation & Reliability

Question 5 6
Customization Ease Very Difficult Easy
Real-world Usage Examples Data fetched through Public Web API
Codebase Quality Poor Good, to be reviewed
Performance & Reliability High performance, Poor reliability due to lack of tests High performance, High reliability, Automated testing : Work-in-progress

  1. In strict technical terms, PVGIS was and remains a database-less application ! ↩↩

  2. https://docs.xarray.dev/en/stable/user-guide/io.html ↩

  3. https://pvlib-python.readthedocs.io/en/stable/reference/generated/pvlib.spa.html ↩