kumulant/com.eignex.kumulant.stat.regression.glm/Link/Logit Logit @Serializable@SerialName(value = "Logit")data object Logit : Link(source)mu = sigmoid(eta). Bernoulli likelihood; expects y in {0, 1}. Members Functions curvature Link copied to clipboard open override fun curvature(eta: Double): DoubleSecond derivative of the per-observation negative log-likelihood at eta. invMean Link copied to clipboard open override fun invMean(eta: Double): DoubleInverse link: maps the linear predictor eta to the response mean. loss Link copied to clipboard open override fun loss(eta: Double, y: Double): Doublesoftplus(eta) - y * eta = log(1 + exp(eta)) - y * eta, computed stably. curvature open override fun curvature(eta: Double): Double(source)Second derivative of the per-observation negative log-likelihood at eta. invMean open override fun invMean(eta: Double): Double(source)Inverse link: maps the linear predictor eta to the response mean. loss open override fun loss(eta: Double, y: Double): Double(source)softplus(eta) - y * eta = log(1 + exp(eta)) - y * eta, computed stably.