R/print.R

In palatej/rjd3highfreq: Seasonal Adjustment of High Frequency Data with 'JDemetra+ 3.x'

#' Print method for 'JDFractionalAirlineDecomposition' objects
#'
#' This function prints informations on the result of a Fractional Airline model (classe JDFractionalAirlineDecomposition).
#'
#' @param x An object of class 'JDFractionalAirlineDecomposition'.
#' @param digits Number of digits to round numerical values (default is 3 or digits - 3 from options).
#'
#' @return The original object 'x'.
#'
#' @export
print.JDFractionalAirlineDecomposition <- function(x, digits = max(3L, getOption("digits") - 3L), ...) {
    cat("Number of observations:", formatC(x$likelihood$nobs, digits = digits))
    cat("\n")

    if (!is.null(x$decomposition$y_time)) {
        cat("Start:", format(x$decomposition$y_time[1]),"\n")
        cat("End:", format(x$decomposition$y_time[length(x$decomposition$y_time)]), "\n")
    }

    # Estimated MA parameters (coefs, se, student)
    nb_freq <- length(x$estimation$parameters) - 1L
    est_ma_params <- data.frame(
        MA_parameter = c("Theta(1)",
                         paste0("Theta(", paste0("period = ",
                                                 x$estimation$periods), ")")),
        Coef = x$estimation$parameters,
        Coef_SE = sqrt(diag(x$estimation$covariance)),
        check.names = FALSE)
    est_ma_params$Tstat <- est_ma_params$Coef / est_ma_params$Coef_SE

    cat("\n")
    cat("Estimate MA parameters:")
    cat("\n")
    print(est_ma_params, row.names = FALSE)
    cat(ifelse(x$estimation$log, "Multiplicative", "Additive"), "model\n")

    cat("\n")
    cat("Decomposition:")
    cat("\n")
    decompo_table <- do.call(cbind, x$decomposition)
    if (x$estimation$log) {
        decompo_table <- exp(decompo_table)
    }

    if (!is.null(x$decomposition$y_time)) {
        decompo_table <- subset(decompo_table, select = -y_time)
        rownames(decompo_table) <- format(x$decomposition$y_time)
    }
    print(tail(decompo_table, n = 10))
    cat("\n")

    cat("Sum of square residuals:", formatC(x$likelihood$ssq, digits = digits),
        "on", x$likelihood$df, "degrees of freedom",
        sep = " ")
    cat("\n")

    cat("Log likelihood = ", formatC(x$likelihood$ll, digits = digits),
        ", \n\taic = ", formatC(x$likelihood$aic, digits = digits),
        ", \n\taicc = ", formatC(x$likelihood$aicc, digits = digits),
        ", \n\tbic(corrected for length) = ",
        formatC(x$likelihood$bicc, digits = digits), sep = "")
    cat("\n")

    cat("Hannan–Quinn information criterion = ",
        formatC(x$likelihood$hannanquinn, digits = digits), sep = "")

    cat("\n\n")

    return(invisible(x))
}

#' Print method for 'JDFractionalAirlineEstimation' objects
#'
#' This function prints informations on the result of a Fractional Airline model (classe JDFractionalAirlineEstimation).
#'
#' @param x An object of class 'JDFractionalAirlineEstimation'.
#' @param digits Number of digits to round numerical values (default is 3 or digits - 3 from options).
#'
#' @return The original object 'x'.
#'
#' @export
print.JDFractionalAirlineEstimation <- function(x, digits = max(3L, getOption("digits") - 3L), ...) {

    cat("Number of observations:", formatC(x$likelihood$nobs, digits = digits))
    cat("\n")

    if (!is.null(x$model$y_time)) {
        cat("Start:", format(x$model$y_time[1]),"\n")
        cat("End:", format(x$model$y_time[length(x$model$y_time)]), "\n")
    }

    nb_outliers <- sum(toupper(substr(x$model$variables, 1L, 2L)) %in%
                           c("AO", "WO", "LS"))
    nb_reg_cjo <- length(x$model$variables) - nb_outliers

    summary_coeff <-  data.frame(
        "Variable" = x$model$variables,
        "Coef" = x$model$b,
        "Coef_SE" = sqrt(diag(x$model$bcov)))
    summary_coeff$Tstat <- round(summary_coeff$Coef / summary_coeff$Coef_SE, digits)
    summary_coeff$Coef <- round(summary_coeff$Coef, digits)
    summary_coeff$Coef_SE <- round(summary_coeff$Coef_SE, digits)

    if (nb_outliers > 0) {
        outliers_coeff <- summary_coeff[(nb_reg_cjo + 1L):nrow(summary_coeff), ]
    }

    if (nb_reg_cjo > 0) {
        reg_cjo_coeff <- summary_coeff[seq_len(nb_reg_cjo), ]
    }

    # Estimated MA parameters (coefs, se, student)
    nb_freq <- length(x$estimation$parameters) - 1L
    est_ma_params <- data.frame(
        MA_parameter = c("Theta(1)",
                         paste0("Theta(period = ", x$model$periods, ")")),
        Coef = x$estimation$parameters,
        Coef_SE = sqrt(diag(x$estimation$covariance)),
        check.names = FALSE
    )
    est_ma_params$Tstat <- est_ma_params$Coef / est_ma_params$Coef_SE

    cat("\n")
    cat("Estimate MA parameters:")
    cat("\n")
    print(est_ma_params, row.names = FALSE)

    cat("\n")
    cat("Number of calendar regressors:", nb_reg_cjo, ", Number of outliers :", nb_outliers)
    cat("\n\n")

    if (nb_reg_cjo > 0) {
        cat("TD regressors coefficients:")
        cat("\n")
        print(reg_cjo_coeff, row.names = FALSE)
        # print(head(reg_cjo_coeff, 10), row.names = FALSE)
        # if (nb_reg_cjo > 10) cat("...\n")
        cat("\n")
    }

    if (nb_outliers > 0) {
        cat("Outliers coefficients:")
        cat("\n")
        print(outliers_coeff, row.names = FALSE)
        # print(head(outliers_coeff, 10), row.names = FALSE)
        # if (nb_outliers > 10) cat("...\n")
        cat("\n")
    }

    cat("Sum of square residuals:", formatC(x$likelihood$ssq, digits = digits),
        "on", x$likelihood$df, "degrees of freedom",
        sep = " ")
    cat("\n")

    cat("Log likelihood = ", formatC(x$likelihood$ll, digits = digits),
        ", \n\taic = ", formatC(x$likelihood$aic, digits = digits),
        ", \n\taicc = ", formatC(x$likelihood$aicc, digits = digits),
        ", \n\tbic(corrected for length) = ",
        formatC(x$likelihood$bicc, digits = digits), sep = "")
    cat("\n")

    cat("Hannan–Quinn information criterion = ",
        formatC(x$likelihood$hannanquinn, digits = digits), sep = "")

    cat("\n\n")
    return(invisible(x))
}