Extract pairwise association weights as a matrix or tidy data frame

Returns the pairwise effect size values from a catgraph object as a symmetric matrix or a long-format tidy data frame suitable for further analysis or export.

Usage

assoc_matrix(
  x,
  format = c("matrix", "tidy"),
  include_p = TRUE,
  include_n = TRUE
)

Arguments

x

A catgraph object.

format

Character. Output format: "matrix" (default) returns a symmetric numeric matrix with NA on the diagonal; "tidy" returns a data frame with one row per pair.

include_p

Logical. When format = "tidy", whether to include the p-value column. Default TRUE.

include_n

Logical. When format = "tidy", whether to include the pairwise observation count. Default TRUE.

Value

format = "matrix"

A symmetric numeric matrix of dimension p x p (p = number of variables). Diagonal is NA. Row and column names are the variable names.

format = "tidy"

A data frame with columns var1, var2, effect_size, metric, type, and optionally p_value and n.

See also

catgraph, prune_edges

Examples

df <- as.data.frame(Titanic)
df_exp <- df[rep(seq_len(nrow(df)), df$Freq), -5]
cg <- catgraph(df_exp)
assoc_matrix(cg)
#>              Class       Sex        Age   Survived
#> Class           NA 0.3987227 0.23194779 0.29412010
#> Sex      0.3987227        NA 0.11101269 0.45560478
#> Age      0.2319478 0.1110127         NA 0.09757511
#> Survived 0.2941201 0.4556048 0.09757511         NA
assoc_matrix(cg, format = "tidy")
#>    var1     var2 effect_size    metric type       p_value    n
#> 5   Sex Survived  0.45560478       phi  2x2 2.302151e-101 2201
#> 1 Class      Sex  0.39872269 cramers_v  RxC  1.556637e-75 2201
#> 3 Class Survived  0.29412010 cramers_v  RxC  4.999928e-41 2201
#> 2 Class      Age  0.23194779 cramers_v  RxC  1.694884e-25 2201
#> 4   Sex      Age  0.11101269       phi  2x2  1.907432e-07 2201
#> 6   Age Survived  0.09757511       phi  2x2  4.700752e-06 2201