zeitzeigerSig: Estimate significance of periodicity by permutation testing.

Description Usage Arguments Value See Also

View source: R/zeitzeiger_fit.R

Description

zeitzeigerSig estimates the statistical significance of the periodic smoothing spline fit. At each permutation, the time vector is scrambled and then zeitzeigerFit is used to fit a periodic smoothing spline for each feature as a function of time. The p-value for each feature is calculated based on the of permutations that had a signal-to-noise ratio at least as large as the observed signal-to-noise ratio, adjusted by the method of Phipson and Smyth (2010). Make sure to first register the parallel backend using registerDoParallel. For genome-scale data, this function will be very slow.

Usage

1
2
zeitzeigerSig(x, time, fitMeanArgs = list(rparm = NA, nknots = 3),
  nIter = 200, dopar = TRUE)

Arguments

x

Matrix of measurements, with observations in rows and features in columns. Missing values are allowed.

time

Vector of values of the periodic variable for the observations, where 0 corresponds to the lowest possible value and 1 corresponds to the highest possible value.

fitMeanArgs

List of arguments to pass to bigspline.

nIter

Number of permutations.

dopar

Logical indicating whether to process features in parallel. Use registerDoParallel to register the parallel backend.

Value

Vector of p-values.

See Also

zeitzeigerFit


jakejh/zeitzeiger documentation built on Sept. 1, 2018, 9:51 a.m.