pyblock API¶
pyblock is a python module for analysis of correlated data.
pyblock.blocking
implements the reblocking algorithm [1] and an
algorithm [2], [3] for suggesting the most appropriate block size (and thus
estimate of the standard error in the data set) for data contained within
numpy
arrays. pyblock.pd_utils
provides a nice wrapper around
this using pandas
, and it is highly recommended to use this if possible.
pyblock.error
contains functions for simple error propagation and
formatting of output of a value and it’s associated error.
References¶
[1] | “Error estimates on averages of correlated data”, H. Flyvbjerg and H.G. Petersen, J. Chem. Phys. 91, 461 (1989). |
[2] | “Monte Carlo errors with less errors”, U. Wolff, Comput. Phys. Commun. 156, 143 (2004) and arXiv:hep-lat/0306017. |
[3] | “Strategies for improving the efficiency of quantum Monte Carlo calculations”, R. M. Lee, G. J. Conduit, N. Nemec, P. Lopez Rios, and N. D. Drummond, Phys. Rev. E. 83, 066706 (2011). |