Methods for data analysis

Locan will provide methods to work on LocData objects and carry out standard analysis procedures. Some of these functions are merely wrapper functions for well established analysis routines in third-party packages.

The methods in this context can either be stand-alone functions with a well-defined input and output and absolutely no side-effects.

In general however, results consist of numeric data, statistical properties (e.g. a histogram of numeric results) and/or fit parameter from comparing the numeric results with theoretical expectations, and annotated plots. In this case the method will be part of a more complex analysis class.

Locan therefore provides specific analysis classes that follow a common structure. Any Analysis object is instantiated with a set of parameters that define the precise analysis procedure. The computation is then performed by calling Analysis.compute() with a parameter specifying one or more LocData objects on which to perform the analysis. The main result is typically provided under the attribute Analysis.results and accompanied by a flexible set of further attributes. Common methods often include plot, hist, and report functions. Metadata is provided under the Analysis.meta attribute.