Skip to contents

Quickly choose an elbow for a PC

quick_elbow.Rd

At variance below 5% per component, choose the largest % drop. Designed for variance percentages, but will also work given a full set of Evalues. Quickly estimate the 'elbow' of a scree plot (PCA)

Usage

quick_elbow(varpc, low = 0.08, max.pc = 0.9)

Arguments

varpc

numeric, vector of eigenvalues, or 'percentage of variance' explained by datapoints for each principal component. If only using a partial set of components, should first pass to estimate.eig.vpcs() to estimate any missing eigenvalues

low

numeric (between zero and one); the threshold that defines whether a principal component explains 'much' of the variance

max.pc

maximum percentage of the variance to capture before the elbow (cumulative sum to PC 'n')

Value

the number of principal components to keep, prior to the determined elbow cutoff

Details

This function uses a rough algorithm to estimate a sensible 'elbow' to choose for a PCA screeplot of eigenvalues. The function looks at an initial arbitrarily 'low' level of variance and looks for the first eigenvalue lower than this. If the very first eigenvalue is actually lower than this (i.e, when the PCs are not very explanatory) then this 'low' value is iteratively halved until this is no longer the case. After starting below this arbitrary threshold the drop in variance explained by each pair of consecutive PCs is standardized by dividing over the larger of the pair. The largest percentage drop in the series below 'low' % is selected as the 'elbow'.

See also

estimate.eig.vpcs

Other LFMM functions: lfmm_df(), lfmm_do_everything(), lfmm_manhattanplot(), lfmm_qqplot(), lfmm_run(), lfmm_table(), lfmm_test_tidy(), select_K()

Author

Nicholas Cooper

Examples

# correlated data
mat <- sim.cor(100, 50)
#> Error in sim.cor(100, 50): could not find function "sim.cor"
result <- princomp(mat)
#> Error in eval(expr, envir, enclos): object 'mat' not found
eig <- result$sdev^2
#> Error in eval(expr, envir, enclos): object 'result' not found
elb.a <- quick_elbow(eig)
#> Error in eval(expr, envir, enclos): object 'eig' not found
pca.scree.plot(eig, elbow = elb.a, M = mat)
#> Error in pca.scree.plot(eig, elbow = elb.a, M = mat): could not find function "pca.scree.plot"
elb.b <- quick_elbow(eig, low = .05) # decrease 'low' to select more components
#> Error in eval(expr, envir, enclos): object 'eig' not found
pca.scree.plot(eig, elbow = elb.b, M = mat)
#> Error in pca.scree.plot(eig, elbow = elb.b, M = mat): could not find function "pca.scree.plot"
# random (largely independent) data, usually higher elbow #
mat2 <- generate.test.matrix(5, 3)
#> Error in generate.test.matrix(5, 3): could not find function "generate.test.matrix"
result2 <- princomp(mat2)
#> Error in eval(expr, envir, enclos): object 'mat2' not found
eig2 <- result2$sdev^2
#> Error in eval(expr, envir, enclos): object 'result2' not found
elb2 <- quick_elbow(result2$sdev^2)
#> Error in eval(expr, envir, enclos): object 'result2' not found
pca.scree.plot(eig2, elbow = elb2, M = mat2)
#> Error in pca.scree.plot(eig2, elbow = elb2, M = mat2): could not find function "pca.scree.plot"