gt_pca_randomSVD: PCA for 'gen_tibble' objects by randomized partial SVD

In tidypopgen: Tidy Population Genetics

PCA for gen_tibble objects by randomized partial SVD

Description

This function performs Principal Component Analysis on a gen_tibble, by randomised partial SVD based on the algorithm in RSpectra (by Yixuan Qiu and Jiali Mei).
This algorithm is linear in time in all dimensions and is very memory-efficient. Thus, it can be used on very large big.matrices. This function is a wrapper for bigstatsr::big_randomSVD()

Usage

gt_pca_randomSVD(
  x,
  k = 10,
  fun_scaling = bigsnpr::snp_scaleBinom(),
  tol = 1e-04,
  verbose = FALSE,
  n_cores = 1,
  fun_prod = bigstatsr::big_prodVec,
  fun_cprod = bigstatsr::big_cprodVec,
  total_var = TRUE
)

Arguments

x	a gen_tibble object
k	Number of singular vectors/values to compute. Default is 10. This algorithm should be used to compute a few singular vectors/values.
fun_scaling	Usually this can be left unset, as it defaults to bigsnpr::snp_scaleBinom(), which is the appropriate function for biallelic SNPs. Alternatively it is possible to use custom function (see bigsnpr::snp_autoSVD() for details.
tol	Precision parameter of svds. Default is 1e-4.
verbose	Should some progress be printed? Default is FALSE.
n_cores	Number of cores used.
fun_prod	Function that takes 6 arguments (in this order): a matrix-like object X, a vector x, a vector of row indices ind.row of X, a vector of column indices ind.col of X, a vector of column centers (corresponding to ind.col), a vector of column scales (corresponding to ind.col), and compute the product of X (subsetted and scaled) with x.
fun_cprod	Same as fun.prod, but for the transpose of X.
total_var	a boolean indicating whether to compute the total variance of the matrix. Default is TRUE. Using FALSE will speed up computation, but the total variance will not be stored in the output (and thus it will not be possible to assign a proportion of variance explained to the components).

Value

a gt_pca object, which is a subclass of bigSVD; this is an S3 list with elements: A named list (an S3 class "big_SVD") of

• d, the eigenvalues (singular values, i.e. as variances),

• u, the scores for each sample on each component (the left singular vectors)

• v, the loadings (the right singular vectors)

• center, the centering vector,

• scale, the scaling vector,

• method, a string defining the method (in this case 'randomSVD'),

• call, the call that generated the object.

Note: rather than accessing these elements directly, it is better to use tidy and augment. See gt_pca_tidiers.

See Also

bigstatsr::big_randomSVD() which this function wraps.

Examples

vcf_path <-
  system.file("extdata", "anolis",
    "punctatus_t70_s10_n46_filtered.recode.vcf.gz",
    package = "tidypopgen"
  )
anole_gt <-
  gen_tibble(vcf_path, quiet = TRUE, backingfile = tempfile("anolis_"))

# Remove monomorphic loci and impute
anole_gt <- anole_gt %>% select_loci_if(loci_maf(genotypes) > 0)
anole_gt <- gt_impute_simple(anole_gt, method = "mode")

# Create PCA object, including total variance
gt_pca_randomSVD(anole_gt, k = 10, total_var = TRUE)

tidypopgen documentation built on Jan. 23, 2026, 1:06 a.m.