pyblock implements the reblocking analysis (see, for example, the description given by Flyvbjerg and Petersen [1]), to remove serial correlation from a data set and hence obtain an improved estimate of the standard error. Functions for additional analysis, interpretation and manipulation of the resultant mean and standard error estimates are also provided.

A command-line interface is currently not provided but the API is simple to use from either a Python/IPython shell or to create an application-specific script.


[1]“Error estimates on averages of correlated data”, H. Flyvbjerg and H.G. Petersen, J. Chem. Phys. 91, 461 (1989).

Indices and tables