ldhmm.calc_stats_from_obs

ldhmm.drop_outliers

This utility computes the statistics (mean, sd, kurtosis, length) for each state.
It can be based on the local or global decoding result. The concept of asymptotic 
statistics can be applied by which the largest N observations (in absolute term)
can be dropped to avoid distortion from outliers. 
It is assumed the object already has come with filled data in 
<code>observations, states.prob, states.local, states.global</code> slots.

mllk

Hidden Markov Model (HMM) based on symmetric lambda distribution
framework is implemented for the study of return time-series in the financial
market. Major features in the S&P500 index, such as regime identification,
volatility clustering, and anti-correlation between return and volatility,
can be extracted from HMM cleanly. Univariate symmetric lambda distribution
is essentially a location-scale family of exponential power distribution.
Such distribution is suitable for describing highly leptokurtic time series
obtained from the financial market. It provides a theoretically solid foundation
to explore such data where the normal distribution is not adequate. The HMM
implementation follows closely the book: "Hidden Markov Models for Time Series",
by Zucchini, MacDonald, Langrock (2016).

Stephen Lihn

ldhmm

Hidden Markov Model for Financial Time-Series Based on Lambda
Distribution

ldhmm.calc_stats_from_obs function

<dl><dt>object</dt>
<dd>an ldhmm object that contains the observations.</dd>
<dt>drop</dt>
<dd>numeric, an integer to drop the largest N observations, default is zero.</dd>
<dt>use.local</dt>
<dd>logical, use local decoding result, default is <code>TURE</code>. 
Otherwise, use global decoding result.</dd>
<dt>x</dt>
<dd>numeric, the observations.</dd></dl>

Arguments

Author

Computing the statistics for each state — ldhmm.calc_stats_from_obs

<dl>

<dt>object</dt>
<dd>an ldhmm object that contains the observations.</dd>


<dt>drop</dt>
<dd>numeric, an integer to drop the largest N observations, default is zero.</dd>


<dt>use.local</dt>
<dd>logical, use local decoding result, default is <code>TURE</code>. 
Otherwise, use global decoding result.</dd>


<dt>x</dt>
<dd>numeric, the observations.</dd>

</dl>

ldhmm.calc_stats_from_obs: Computing the statistics for each state

Description

Usage

Value

Arguments

Author