Minimum significant correlation for a sample size

Description

minSigCor is a helper function that estimates the minimum significant correlation for a smaple size n at a confidence level defined by the argument alpha.

Usage

1
minSigCor(n=41, alpha=0.05, r=seq(0, 1, by=1e-6))

Arguments

n

sample size or the length of a timeseries vector.

alpha

confidence level: the default is alpha = 0.05 for 95% confidence level.

r

a vector of values from 0 to 1 to search for the minimum signifcant correlation for the user-specified sample size n at confidence level alpha. This should be a subset of the valid positiva correlation range 0-1. The default is to search for the minimum significant correlation in the complete range 0-1 with a very fine step of 1e-6. For faster computations, the user may set a shorter range with larger step (e.g., seq(0.1, 0.5, by=1e-3)).

Details

minSigCor function estimates the minimum significant correlation for a sample size (number of observations or temporal points in a timeseries) at a certain confidence level selected by the argument alpha and an optional search range r. It is called by validClimR function objective tree cut based on the specified confidence level.

Value

A positive value beween 0 and 1 for the estimated the minimum significant correlation.

Author(s)

Hamada Badr <badr@jhu.edu>, Ben Zaitchik <zaitchik@jhu.edu>, and Amin Dezfuli <dez@jhu.edu>.

References

Hamada S. Badr, Zaitchik, B. F. and Dezfuli, A. K. (2015): A Tool for Hierarchical Climate Regionalization, Earth Science Informatics, 1-10, http://dx.doi.org/10.1007/s12145-015-0221-7.

Hamada S. Badr, Zaitchik, B. F. and Dezfuli, A. K. (2014): Hierarchical Climate Regionalization, CRAN, http://cran.r-project.org/package=HiClimR.

See Also

HiClimR, validClimR, geogMask, coarseR, fastCor, grid2D, and minSigCor.

Examples

1
2
3
4
require(HiClimR)

## Find minimum significant correlation at 95% confidence level
rMin <- minSigCor(n = 41, alpha = 0.05, r = seq(0, 1, by = 1e-06))