QExPy (Queen’s Experimental Physics) is a python 3 package designed to facilitate data analysis in undergraduate physics laboratories. The package contains a module to easily propagate errors in uncertainty calculations, and a module that provides an intuitive interface to plot and fit data. The package is designed to be efficient, correct, and to allow for a pedagogic introduction to error analysis. The package is extensively tested in the Jupyter Notebook environment to allow high quality reports to be generated directly from a browser.
Highlights:
- Easily propagate uncertainties in measured quantities
- Compare different uncertainty calculations (e.g. Min-Max, quadrature errors, Monte Carlo errors)
- Correctly include correlations between quantities when propagating uncertainties (e.g. the uncertainty on x-x should always be 0!)
- Calculate exact numerical values of derivatives
- Choose display format (standard, Latex, scientific notation)
- Control the number of significant figures
- Handle ensembles of measurements (e.g. combine multiple measurements, with uncertainties, of a single quantity)
- Produce interactive plots of data in the browser
- Fit data to common functions (polynomials, gaussians) or provide custom functions
- Examine residual plots after fits
- Track units in calculations (still in development)
- Plot confidence bands from the errors in fitted parameters (still in development)
- Integrates with Jupyter notebooks, numpy, bokeh
Up to date examples are maintained in the examples directory of the repository. These are likely the best way to get acquainted with the package.
Refer to the example notebooks in the examples/jupyter directory to learn how to use the package, and browse through the official documentstion.
Read the documentation at http://qexpy.readthedocs.io/en/latest/intro.html