# Panel Estimators for Generalized Linear Models Estimation by maximum likelihood of glm (binomial and Poisson) and 'glm-like' models (Negbin and ordered) on longitudinal data ```r pglm(formula, data, subset, na.action, effect = c("individual", "time", "twoways"), model = c("random", "pooling", "within", "between"), family, other = NULL, index = NULL, start = NULL, R = 20, ...) ``` ## Arguments - `formula`: a symbolic description of the model to be estimated, - `data`: the data: a `pdata.frame` object or an ordinary `data.frame`, - `subset`: an optional vector specifying a subset of observations, - `na.action`: a function which indicates what should happen when the data contains '`NA`'s, - `effect`: the effects introduced in the model, one of `"individual"`, `"time"` or `"twoways"`, - `model`: one of `"pooling"`, `"within"`, `"between"`, `"random",`, - `family`: the distribution to be used, - `other`: for developper's use only, - `index`: the index, - `start`: a vector of starting values, - `R`: the number of function evaluation for the gaussian quadrature method used, - `...`: further arguments. ## Returns An object of class `"pglm"`, a list with elements: - **coefficients**: the named vector of coefficients, - **logLik**: the value of the log-likelihood, - **hessian**: the hessian of the log-likelihood at convergence, - **gradient**: the gradient of the log-likelihood at convergence, - **call**: the matched call, - **est.stat**: some information about the estimation (time used, optimisation method), - **freq**: the frequency of choice, - **residuals**: the residuals, - **fitted.values**: the fitted values, - **formula**: the formula (a `mFormula` object), - **expanded.formula**: the formula (a `formula` object), - **model**: the model frame used, - **index**: the index of the choice and of the alternatives. ## Examples ```r ## an ordered probit example data('Fairness', package = 'pglm') Parking <- subset(Fairness, good == 'parking') op <- pglm(as.numeric(answer) ~ education + rule, Parking[1:105, ], family = ordinal('probit'), R = 5, print.level = 3, method = 'bfgs', index = 'id', model = "random") ## a binomial (probit) example data('UnionWage', package = 'pglm') anb <- pglm(union ~ wage + exper + rural, UnionWage, family = binomial('probit'), model = "pooling", method = "bfgs", print.level = 3, R = 5) ## a gaussian example on unbalanced panel data data(Hedonic, package = "plm") ra <- pglm(mv ~ crim + zn + indus + nox + age + rm, Hedonic, family = gaussian, model = "random", print.level = 3, method = "nr", index = "townid") ## some count data models data("PatentsRDUS", package="pglm") la <- pglm(patents ~ lag(log(rd), 0:5) + scisect + log(capital72) + factor(year), PatentsRDUS, family = negbin, model = "within", print.level = 3, method = "nr", index = c('cusip', 'year')) la <- pglm(patents ~ lag(log(rd), 0:5) + scisect + log(capital72) + factor(year), PatentsRDUS, family = poisson, model = "pooling", index = c("cusip", "year"), print.level = 0, method="nr") ## a tobit example data("HealthIns", package="pglm") HealthIns$med2 <- HealthIns$med / 1000 HealthIns2 <- HealthIns[-2209, ] set.seed(2) subs <- sample(1:20186, 200, replace = FALSE) HealthIns2 <- HealthIns2[subs, ] la <- pglm(med ~ mdu + disease + age, HealthIns2, model = 'random', family = 'tobit', print.level = 0, method = 'nr', R = 5) ``` ## Author(s) Yves Croissant

pglm function