Mean and standard deviation of correlation

For the given data matrix, the mean and standard deviation of correlation are calculated for every pair of variables. The mean and standard deviation are calculated based on correlation values obtained after random subsampling.

cor_resamplingStats(
  data,
  size_of_subsample,
  replace = FALSE,
  n_repetitions = 100,
  zero_precisionC = 1e-06
)

Arguments

data	Data matrix or data frame with samples in rows and variables in columns.
size_of_subsample	Size of subsample taken in every repetition.
replace	Defines whether the resampling is done with replacement. Defaults to `FALSE`.
n_repetitions	Number of repetitions of subsampling. For large `data` matrices, large number of repetitions can take a long time.
zero_precisionC	The zero precision passed to the C++ function. All numbers with absolute value smaller than `zero_precisionC` will be set to zero. Defaults to 10^(-6).

Value

A data frame with names of variables in first two columns. In third columns is mean and in the fourth column is standard deviation for the given pair of variables. In the rows, all pairwise correlations are listed.

Details

For `n_repetitions` times, a random subsample is taken based on parameters `size_of_subsample` and `replace`. Then for every pair of variables a correlation is calculated for a random subsample. At the end of repetitions, the mean and standard deviation of correlation are calculated for every pair of variables.