Tidy a correlation matrix — tidy.cor

This function turns the correlation matrix returned by cor_matrix() and its relatives into a tidy dataframe. Note that by default, results for both the cor(A, B) and cor(B, A) are returned, while entries for A, A, i.e. the values on the diagonal, are never included.

# S3 method for class 'cor_matrix'
tidy(x, both_directions = TRUE, ...)

Arguments

x: A cor_marix object returned from cor_matrix()
both_directions: Should both cor(A, B) and cor(B, A) be returned. Defaults to TRUE.
...: Additional arguments. Not used. Needed to match generic signature only.

Value

A tibble::tibble() with columns:

column1: Name of the first variable
column2: Name of the second variable
estimate: The estimated value of the correlation
statistic: The t-statistic used for significance testing
p.value: The two-sided p-value of the correlation
n: Number of observations used to compute the correlation
ci.low: Lower bound of confidence interval. Width is determined in call to cor_matrix()
ci.high: Upper bound of confidence interval. Width is determined in call to cor_matrix()

Examples

# Create and tidy a correlation matrix
cm <- cor_matrix(iris)
tidy(cm, both_directions = FALSE)
#> # A tibble: 6 × 9
#>   column1 column2 estimate std.error  p.value statistic     n conf.low conf.high
#>   <chr>   <chr>      <dbl>     <dbl>    <dbl>     <dbl> <dbl>    <dbl>     <dbl>
#> 1 Sepal.… Sepal.…   -0.118    0.0816 1.52e- 1     -1.44   150   -0.273    0.0435
#> 2 Petal.… Sepal.…    0.872    0.0403 1.04e-47     21.6    150    0.827    0.906 
#> 3 Petal.… Sepal.…    0.818    0.0473 2.33e-37     17.3    150    0.757    0.865 
#> 4 Petal.… Sepal.…   -0.428    0.0743 4.51e- 8     -5.77   150   -0.551   -0.288 
#> 5 Petal.… Sepal.…   -0.366    0.0765 4.07e- 6     -4.79   150   -0.497   -0.219 
#> 6 Petal.… Petal.…    0.963    0.0222 4.68e-86     43.4    150    0.949    0.973 

# With ESS health data
cm_ess <- cor_matrix(ess_health,
                     var_names = c(health = "Health", dosprt = "Sport"))
tidy(cm_ess)
#> # A tibble: 2 × 9
#>   column2 column1 estimate std.error  p.value statistic     n conf.low conf.high
#>   <chr>   <chr>      <dbl>     <dbl>    <dbl>     <dbl> <dbl>    <dbl>     <dbl>
#> 1 Sport   Health    -0.183    0.0116 5.86e-55     -15.7  7197   -0.205    -0.160
#> 2 Health  Sport     -0.183    0.0116 5.86e-55     -15.7  7197   -0.205    -0.160