Search

Reproducibility is a key aspect of science, and both, publishers and funders increasingly ask for the raw data and details of the data processing to be available. However, still most papers lack the details necessary to fully reproduce the data analysis. This can in part be attributed to a lack of appropriate software tools helping the researchers to document both, data acquisition and analysis automatically and with the required details. We present a Python framework for the fully reproducible Analysis of Spectroscopic Data (AspecD) requiring no programming skills of the users thanks to a unique user interface in form of recipes formatted as YAML file. Based on this framework, a number of packages has been developed, most notably for the analysis of cw-EPR and tr-EPR data. Furthermore, we show a simple yet powerful file format for recording metadata during data acquisition that is both, human-writable and machine-readable. Together, these software tools tremendously enhance the reproducibility of our data analysis and hence the overall quality of the reported science.

As Newton famously stated, we’re all standing on the shoulders of giants: The progress of science rests on the published work of others. In the digital age, we are confronted with an exponential growth of available data. Hence, appropriate research data management is urgently needed and data need to be findable, accessible, interoperable, and reusable (FAIR), prompting initiatives such as the German NFDI. However, a necessary though often underrepresented prerequisite for FAIR data is a gapless record of their provenance. We therefore need tools that automatically record each step from the raw/primary data to their final shareable representation and take care of the respective metadata of each step starting with the data acquisition. Here, we present both, a general framework for the reproducible analysis of spectral data (ASpecD) as well as concrete packages for working with continuous-wave and time-resolved EPR data. Key features are an automatic and gapless record of each processing and analysis step as well as an intuitive yet powerful user interface requiring no programming skills. Additional packages actively being developed demonstrate the potential of the underlying ASpecD framework, dealing with optical absorption and NMR data. Furthermore, the FitPy package allows for advanced fitting of models to spectroscopic data. Together, these tools fill a crucial gap. In context of a larger infrastructure for research data management – including an electronic lab notebook (ELN), wiki-based knowledge management and data stores providing unique IDs, i.e. a laboratory information and management system (LIMS) – they allow for actually thinking about FAIRer EPR (and other spectroscopic) data.