callback (version 0.1.3)

reg_als: Asymptotic least squares estimation

Description

Asymptotic least squares estimation

Usage

reg_als(x, y, omega, ols = FALSE)

Value

a list with class reg_als containing "config"

for the definition of the estimation method and "reg" for the estimation output.

The "config" data frame includes the following elements:

• family: "als" (for Asymptotic least squares).

• method: "ols" for Ordinary least square, or "fgls" for Feasible generalized least squares.

The list "reg" includes the following elements (when relevant):

• estim:a data frame with c_names for the component names, coef, the estimated coefficients, std_coef, the estimated standard errors, studentthe Student statistics for the equality of the coefficient to 0, p_value, the p-values of the asymptotic Student test.

• cova: the estimated covariance matrix of the estimator.

• over_test:a data frame with the output of the overidentification test (FGLS only). The statistic is given by stat, the degrees of freedom by df and the p-value by p_value

Arguments

x

the matrix of the right-hand variables (incl. the constant term when needed).

y

the vector of the left-hand variable.

omega

the covariance matrix of the disturbances.

ols

logical indicating whether to perform OLS (TRUE) or FGLS (FALSE). The default is FALSE.

References

Chamberlain, G. (1982). Multivariate regression models for panel data. Journal of econometrics, 18(1), 5-46. Gourieroux, C., Monfort, A., & Trognon, A. (1985). Moindres carrés asymptotiques. Annales de l'INSEE, 91-122. Kodde, D. A., Plam, F. C., & Pfann, G. A. (1990). Asymptotic least‐squares estimation efficiency considerations and applications. Journal of Applied Econometrics, 5(3), 229-243.

Examples

model <- list(c("license"),c("woman"),c("woman","license","inter"))
comp <- callback_comp(mobility1,"offer",c("gender","licenses"),"callback",model)
    x <- comp$aux_boole
    y <- comp$aux_coef
omega <- comp$aux_vcov
 
str(reg_als(x,y,omega))