Resampling Collapsed Null Distributions for Simultaneous Inference


[Up] [Top]

Documentation for package ‘reconsi’ version 1.2.0

Help Pages

calcWeights Obtain weights as posterior probabilities to calculate the consensus null
estNormal Fast estimation of mean and standard deviation of a normal distrbution, optionally with weights
estP0 Estimate the fraction of true null hypotheses.
getApproxCovar Obtain a null covariance matrix of binned test statistics
getFdr Calculate tail-area (Fdr) and local (fdr) false discovery rates, based on a certain null distribution
getG0 Obtain the consensus null
getTestStats A function to calculate observed and permuation z-statistics on a n-by-p matrix of observations
getTstat A function to obtain a t-test statistic efficiently. For internal use only
plotApproxCovar Plot an approximatio of the correlation structure of the test statistics
plotNull Plot the obtained null distribution along with a histogram of observed test statistics
ptEdit A custom function to calculate the distribution function of the t-test statistic. For internal use only
qtEdit A custom function to calculate the quantile function of the t-test statistic. For internal use only
quantCorrect Correct quantiles by not returning 0 or 1
reconsi Perform simultaneous inference through collapsed resampling null distributions
rowMultiply A function to efficiently row multiply a a-by-b matrix by a vector of length b. More memory intensive but that does not matter with given matrix sizes
stabExp A function to numerically stabilize an exponentiation. For internal use only
testDAA A function to test for differential absolute abundance on a phyloseq object
testDAA-method A function to test for differential absolute abundance on a phyloseq object
Vandeputte Microbiomes of Crohn's disease patients and healthy controls