Unified Causal Model Summary — causal

Returns a clean one-row data frame with the key causal estimate, standard error, confidence interval, p-value, number of observations, and outcome variable name. This is useful for programmatic comparison of many models.

Usage

causal_summary(model, method_name = NULL, digits = 3, conf_level = 0.95, ...)

Arguments

model: A fitted model object. Supported: lm, glm, fixest, ivreg, and any model class with a summary() method that returns a coefficient table.
method_name: Character or NULL. Label for the estimation method. If NULL, the model class is used.
digits: Integer. Number of decimal places for display. Default 3.
conf_level: Numeric. Confidence level for the interval. Default 0.95.
...: Additional arguments passed to broom::tidy() if available.

Value

An object of class "causal_summary" inheriting from "data.frame". Columns: method, term, estimate, std_error, ci_lower, ci_upper, p_value, n_obs, outcome.

Details

One-Line Tidy Summary for Any Causal Model

Extracts a standardised single-row data frame from any causal inference model (lm, glm, fixest, ivreg, etc.). Works via broom if available, with manual fallback extraction for common model classes. A custom S3 print method provides a clean, console-friendly summary.

Examples

set.seed(1)
n  <- 300
x  <- rnorm(n)
y  <- 2 * x + rnorm(n)
df <- data.frame(y = y, x = x)

m_lm <- lm(y ~ x, data = df)
cs   <- causal_summary(m_lm, method_name = "OLS")
print(cs)
#> 
#> === Causal Summary ===
#>   Method  : OLS
#>   Outcome : y
#>   Term    : x
#>   Estimate: 2.026  (SE: 0.063)
#>   95% CI  : [1.902, 2.149]
#>   p-value : 0.0000
#>   N obs   : 300
#> ======================
#> 

if (requireNamespace("fixest", quietly = TRUE)) {
  m_fe <- fixest::feols(y ~ x, data = df)
  cs2  <- causal_summary(m_fe, method_name = "OLS (fixest)")
  print(cs2)
}
#> 
#> === Causal Summary ===
#>   Method  : OLS (fixest)
#>   Outcome : y
#>   Term    : x
#>   Estimate: 2.026  (SE: 0.063)
#>   95% CI  : [1.902, 2.149]
#>   p-value : 0.0000
#>   N obs   : 300
#> ======================
#>