Correlations

• cor_partial_spearman(): partial rank correlation: Spearman correlation between x and y after removing (linear regression) the effect of the covariates Z on the ranks of x and y; t test with n - 2 - k degrees of freedom (k = number of covariate columns). Without covariates it is the ordinary Spearman correlation.

• cor_matrix_test(): correlation, p-value, adjusted p-value and N matrices on pairwise complete observations.

Usage

cor_partial_spearman(x, y, Z = NULL)

cor_matrix_test(data, vars, method = "spearman", p_adjust = "BH")

Arguments

x, y

Numeric vectors.

Z

Optional data frame of covariates (same number of rows).

data

A data frame.

vars

Numeric variables (names or tidyselect).

method

Correlation method of stats::cor.test().

p_adjust

Multiplicity correction (stats::p.adjust() method).

Value

cor_partial_spearman(): list(estimate, p.value, n). cor_matrix_test(): list(r, p, p_adj, n, long, method, p_adjust), long holding one row per pair.

Examples

cor_partial_spearman(cohort_sim$NFL_V0, cohort_sim$MOCA_V0, cohort_sim[, c("AGE", "SEX")])
#> $estimate
#> [1] -0.5133259
#> 
#> $p.value
#> [1] 2.783139e-09
#> 
#> $n
#> [1] 120
#> 
cm <- cor_matrix_test(cohort_sim, c(NFL_V0, IL6_V0, CRP_V0, MOCA_V0, UPDRS_V0))
round(cm$r, 2)
#>          NFL_V0 IL6_V0 CRP_V0 MOCA_V0 UPDRS_V0
#> NFL_V0     1.00   0.06  -0.06   -0.55     0.40
#> IL6_V0     0.06   1.00   0.05   -0.05     0.04
#> CRP_V0    -0.06   0.05   1.00   -0.07    -0.11
#> MOCA_V0   -0.55  -0.05  -0.07    1.00    -0.20
#> UPDRS_V0   0.40   0.04  -0.11   -0.20     1.00