fit_ACF_monthly

Wraps fit_ACF to fit ACF models separately to each calendar-month sub-series of an xts object. The input series is split by calendar month, fit_ACF is applied to each resulting sub-series, and the fitted parameters are assembled into a parameter matrix with months as rows. A 12-panel patchwork plot of fitted ACF curves is produced, with empty panels for months absent from the data. The grid layout is controlled by nrow and ncol. The three model types (CAS, HK, SRD) supported by fit_ACF are available and may be specified singly or in combination.

Usage

fit_ACF_monthly(ts, lag, type = list("CAS", "HK", "SRD"), nrow = 4, ncol = 3)

Arguments

ts

An xts object containing the time series data.

lag

Integer. Maximum lag passed to fit_ACF.

type

Character vector of ACF model types. Options are "CAS", "HK", and "SRD". Default all three.

nrow

Integer. Number of rows in the panel grid. Default 4.

ncol

Integer. Number of columns in the panel grid. Default 3.

Value

A list with elements ACF_params_monthly (matrix of fitted parameters with months as rows) and ACF_monthly_plot (12-panel patchwork of fitted ACF ggplot objects with a shared legend).

Examples

ts <- xts::xts(rnorm(365), order.by = as.Date("2020-01-01") + 0:364)
monthly_acfs <- fit_ACF_monthly(ts, lag = 10)
monthly_acfs$ACF_monthly_plot