init_kf

<code>init_kf</code> computes the starting values x0 and P0 by generalized least
squares using the first n observations.

Build customized transfer function and ARIMA models with multiple operators
and parameter restrictions. Provides tools for model identification, estimation
using exact or conditional maximum likelihood, diagnostic checking, automatic outlier
detection, calendar effects, forecasting, and seasonal adjustment. The new version
also supports unobserved component ARIMA model specification and estimation
for structural time series analysis.

Jose L. Gallego

tfarima

Transfer Function and ARIMA Models

init_kf function

<dl><dt>mdl</dt>
<dd>an object of class <code>ssm</code>.</dd>
<dt>z</dt>
<dd>optional time series if it differs from model series.</dd>
<dt>n</dt>
<dd>integer, number of observations used to estimate the initial
conditions. If n &lt; d (dimension of state vector), it defaults to length(z).</dd></dl>

Arguments

Initialization of Kalman filter — init_kf

<dl>

<dt>mdl</dt>
<dd>an object of class <code>ssm</code>.</dd>


<dt>z</dt>
<dd>optional time series if it differs from model series.</dd>


<dt>n</dt>
<dd>integer, number of observations used to estimate the initial
conditions. If n &lt; d (dimension of state vector), it defaults to length(z).</dd>

</dl>

init_kf: Initialization of Kalman filter

Description

Usage

Value

Arguments