Pairwise correlations among numeric columns with test statistics,
p-values and (Pearson) confidence intervals. Each pair uses its own
pairwise-complete rows (rows where both variables are observed, the
stats::cor.test() convention); n_used is the per-pair complete count.
Usage
stat_cor(
data,
.cols,
method = c("pearson", "spearman", "kendall"),
.by = NULL,
alternative = "two.sided",
conf_level = 0.95,
p_adjust = "none"
)Arguments
- data
A data frame.
- .cols
<
tidy-select> Numeric columns (at least 2).- method
"pearson","spearman", or"kendall". Confidence intervals are only available for Pearson (NAotherwise).- .by
<
tidy-select> Optional slice columns.- alternative
"two.sided","greater", or"less".- conf_level
Confidence level (Pearson only).
- p_adjust
P-value adjustment across returned rows.
Value
A tibble with class stat_infer: .by identifiers, var1,
var2, n_used, estimate, statistic, df, p.value,
p.adjusted, sig, conf_low, conf_high, method. Each row is
computed on the pairwise-complete rows of its two variables
(n_used = that pair's complete-row count).
Examples
stat_cor(mtcars, .cols = mpg:wt)
#> Correlation test (pearson, two.sided)
#> # A tibble: 15 × 12
#> var1 var2 n_used estimate statistic df p.value conf_low conf_high
#> * <chr> <chr> <int> <dbl> <dbl> <int> <dbl> <dbl> <dbl>
#> 1 mpg cyl 32 -0.852 -8.92 30 6.11e-10 -0.926 -0.716
#> 2 mpg disp 32 -0.848 -8.75 30 9.38e-10 -0.923 -0.708
#> 3 mpg hp 32 -0.776 -6.74 30 1.79e- 7 -0.885 -0.586
#> 4 mpg drat 32 0.681 5.10 30 1.78e- 5 0.436 0.832
#> 5 mpg wt 32 -0.868 -9.56 30 1.29e-10 -0.934 -0.744
#> 6 cyl disp 32 0.902 11.4 30 1.80e-12 0.807 0.951
#> 7 cyl hp 32 0.832 8.23 30 3.48e- 9 0.682 0.915
#> 8 cyl drat 32 -0.700 -5.37 30 8.24e- 6 -0.843 -0.465
#> 9 cyl wt 32 0.782 6.88 30 1.22e- 7 0.597 0.889
#> 10 disp hp 32 0.791 7.08 30 7.14e- 8 0.611 0.893
#> 11 disp drat 32 -0.710 -5.53 30 5.28e- 6 -0.849 -0.481
#> 12 disp wt 32 0.888 10.6 30 1.22e-11 0.781 0.944
#> 13 hp drat 32 -0.449 -2.75 30 9.99e- 3 -0.690 -0.119
#> 14 hp wt 32 0.659 4.80 30 4.15e- 5 0.403 0.819
#> 15 drat wt 32 -0.712 -5.56 30 4.78e- 6 -0.850 -0.484
#> # ℹ 3 more variables: method <chr>, p.adjusted <dbl>, sig <chr>
stat_cor(mtcars, .cols = mpg:wt, method = "spearman")
#> Warning: cannot compute exact p-value with ties
#> Warning: cannot compute exact p-value with ties
#> Warning: cannot compute exact p-value with ties
#> Warning: cannot compute exact p-value with ties
#> Warning: cannot compute exact p-value with ties
#> Warning: cannot compute exact p-value with ties
#> Warning: cannot compute exact p-value with ties
#> Warning: cannot compute exact p-value with ties
#> Warning: cannot compute exact p-value with ties
#> Warning: cannot compute exact p-value with ties
#> Warning: cannot compute exact p-value with ties
#> Warning: cannot compute exact p-value with ties
#> Warning: cannot compute exact p-value with ties
#> Warning: cannot compute exact p-value with ties
#> Warning: cannot compute exact p-value with ties
#> Correlation test (spearman, two.sided)
#> # A tibble: 15 × 12
#> var1 var2 n_used estimate statistic df p.value conf_low conf_high
#> * <chr> <chr> <int> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 mpg cyl 32 -0.911 10425. NA 4.69e-13 NA NA
#> 2 mpg disp 32 -0.909 10415. NA 6.37e-13 NA NA
#> 3 mpg hp 32 -0.895 10337. NA 5.09e-12 NA NA
#> 4 mpg drat 32 0.651 1902. NA 5.38e- 5 NA NA
#> 5 mpg wt 32 -0.886 10292. NA 1.49e-11 NA NA
#> 6 cyl disp 32 0.928 395. NA 2.28e-14 NA NA
#> 7 cyl hp 32 0.902 536. NA 1.87e-12 NA NA
#> 8 cyl drat 32 -0.679 9160. NA 1.94e- 5 NA NA
#> 9 cyl wt 32 0.858 776. NA 3.57e-10 NA NA
#> 10 disp hp 32 0.851 813. NA 6.79e-10 NA NA
#> 11 disp drat 32 -0.684 9186. NA 1.61e- 5 NA NA
#> 12 disp wt 32 0.898 558. NA 3.35e-12 NA NA
#> 13 hp drat 32 -0.520 8294. NA 2.28e- 3 NA NA
#> 14 hp wt 32 0.775 1229. NA 1.95e- 7 NA NA
#> 15 drat wt 32 -0.750 9550. NA 7.59e- 7 NA NA
#> # ℹ 3 more variables: method <chr>, p.adjusted <dbl>, sig <chr>