Summary method for modality_gravity output

Returns a compact role-distribution table and per-variable breakdown, and prints a readable synopsis to the console.

Usage

# S3 method for class 'modality_gravity'
summary(object, ...)

Arguments

object

A data frame returned by modality_gravity.

...

Ignored.

Value

Invisibly returns a list with elements:

role_counts

Named integer vector of attractor/neutral/satellite counts.

by_variable

Data frame with one row per variable showing the dominant role and mean delta_mgi across its modalities.

top_attractor

Single-row data frame: modality with highest delta_mgi.

top_satellite

Single-row data frame: modality with lowest delta_mgi.

spearman_strength

Spearman correlation between strength and delta_mgi.

See also

modality_gravity, print.modality_gravity

Examples

data(survey_health)
mg   <- build_modality_graph(survey_health)
mg   <- prune_modality_edges(mg, min_weight = 0.10, max_p = 0.05)
grav <- modality_gravity(mg)
summary(grav)
#> 
#> === Modality Gravity - Summary ===
#> 
#> Role distribution:
#> 
#> attractor   neutral satellite 
#>        12         2         8 
#> 
#> Dominant attractor : lung_disease=no  (dMGI = 1.958, prev = 0.855)
#> Strongest satellite: bmi_category=obese  (dMGI = -0.508, prev = 0.224)
#> 
#> Spearman rho (strength vs dMGI): 0.191
#> (Values far from +/-1.0 indicate MGI captures structure beyond connectivity)
#> 
#> Per-variable breakdown:
#>          variable n_modalities mean_delta dominant_role
#>      lung_disease            2      0.813     attractor
#>    smoking_status            3      0.658     attractor
#>  health_insurance            2      0.467     attractor
#>               sex            2      0.368     attractor
#>     exercise_freq            3      0.302     attractor
#>         age_group            3      0.219     satellite
#>      bmi_category            4      0.120       neutral
#>      diet_quality            3     -0.002     satellite