Calculate Covariance Matrix in Parallel

Speed up covariance calculation for large matrices. The default behavior is similar cov. Please remove any NA prior to calculation.

Usage

fastcov2(x, y = NULL, col1, col2, df)

Arguments

x

a numeric vector, matrix or data frame; a matrix is highly recommended to maximize the performance

y

NULL (default) or a vector, matrix or data frame with compatible dimensions to x; the default is equivalent to y = x

col1

integers indicating the subset (columns) of x to calculate the covariance; default is all the columns

col2

integers indicating the subset (columns) of y to calculate the covariance; default is all the columns

df

a scalar indicating the degrees of freedom; default is nrow(x)-1

Value

A covariance matrix of x and y. Note that there is no NA handling. Any missing values will lead to NA in the resulting covariance matrices.

Examples

x <- matrix(rnorm(400), nrow = 100)

# Call `cov(x)` to compare
fastcov2(x)
#>             [,1]       [,2]        [,3]       [,4]
#> [1,]  1.17251985 -0.1463312  0.08239227  0.1336536
#> [2,] -0.14633119  1.1509356 -0.12009329 -0.0259273
#> [3,]  0.08239227 -0.1200933  0.91439945  0.0385713
#> [4,]  0.13365361 -0.0259273  0.03857130  0.8697899

# Calculate covariance of subsets
fastcov2(x, col1 = 1, col2 = 1:2)
#>         [,1]       [,2]
#> [1,] 1.17252 -0.1463312

# Speed comparison
x <- matrix(rnorm(100000), nrow = 1000)
microbenchmark::microbenchmark(
  fastcov2 = {
    fastcov2(x, col1 = 1:50, col2 = 51:100)
  },
  cov = {
    cov(x[,1:50], x[,51:100])
  },
  unit = 'ms', times = 10
)
#> Unit: milliseconds
#>      expr      min       lq     mean   median       uq      max neval
#>  fastcov2 1.411794 1.431020 1.532678 1.493262 1.597523 1.833020    10
#>       cov 5.840807 5.866956 6.038163 5.938975 6.177085 6.431409    10