Jupyter notebooks demonstrating the bootstrap common mean direction test (CMDT) proposed in:
D. Heslop, J. L. Scealy, A. T. A. Wood, L. Tauxe and A. P. Roberts (2023). A bootstrap common mean direction test. Journal of Geophysical Research.
Specifically:
Bootstrap confidence region.ipynb
: Estimate the confidence region of the mean of a set of directions.
Common mean direction test.ipynb
: Test if two sets of directions share a common mean.
The Python libraries numpy
, pandas
, matplotlib
, and pmagpy
are needed for the bootstrap-CMDT procedure.
Data files are provided for the examples. Please note, you will need to specify the path of the data files if they are not stored in the same location as the notebooks.
Bootstrapping is a numerical procedure and therefore results may differ slightly from those in the paper.