mrs_plot.m

Modeling and Forecasting Electricity Loads and Prices: A Statistical Approach" by Rafa&#322 Weron, p

function mrs_plot(PrSpike,Data)
%MRS_PLOT Plot calibration results for a 2-state MRS model.
%   MRS_PLOT(PRSPIKE,DATA) generates a figure with two panels visualizing
%   calibration results for a 2-state (Markov) regime switching model 
%   fitted to time series DATA. In the bottom panel probability PRSPIKE 
%   that an observation comes from the spike regime is plotted. In the top 
%   panel the time series is displayed with observations identified as 
%   belonging to the spike regime (i.e. having PRSPIKE>0.5) plotted in red.  
%
%   See also MRS_EST, MRS_SIM.
%
%   Reference(s):
%   [1] R.Weron (2007) 'Modeling and Forecasting Electricity Loads and 
%   Prices: A Statistical Approach', Wiley, Chichester.   

%   Written by Jakub Jurdziak and Rafal Weron (2006.09.21)
%   Copyright (c) 2006 by Rafal Weron

% Define spike regime, non-spiky data and spiky data
SpikeInd = PrSpike > 0.5;
NonSpikyData = Data;
NonSpikyData(find(SpikeInd == 1)) = nan;
SpikeInd = min(length(Data),max(1,[find(SpikeInd == 1); find(SpikeInd == 1)+1; find(SpikeInd == 1)-1]));
SpikyData = nan*Data;
SpikyData(SpikeInd) = Data(SpikeInd);

% Plot data with red spikes
subplot(2,1,1)
plot(1:length(Data), SpikyData, 'red', 1:length(Data), NonSpikyData, 'black');
legend('spike', 'base');
ylabel('Data');
xlim([1 length(Data)]);

% Plot the probability of being in the spike regime
subplot(2,1,2);
plot(1:length(Data), PrSpike);
ylabel('Pr(spike)');
xlim([1 length(Data)]);
ylim([0 1]);