fixed spelling mistake + misc

ddToolbox/classes/ModelBaseClass.m

@@ -324,7 +324,6 @@ function figParticipant(obj, pSamples, pData, mMEAN, cMEAN, epsilonMEAN, alphaME
 		function figUnivariateSummary(obj, participantIDlist, variables)
 			% loop over variables provided, plotting univariate summary
 			% statistics.
-			warning('Add group-level inferences to this plot (glM, glC, glALPHA,glEpsilon)')
 			figure
 			for v = 1:numel(variables)
 				subplot(numel(variables),1,v)

ddToolbox/classes/ModelSeperate.m → ddToolbox/classes/ModelSeparate.m

@@ -1,4 +1,4 @@
-classdef ModelSeperate < ModelBaseClass
+classdef ModelSeparate < ModelBaseClass
 	%ModelSeperate A model to estimate the magnitide effect.
 	%	Models a number of participants, but they are all treated as independent.
 	%	There is no group-level estimation.
@@ -10,14 +10,14 @@
 	methods (Access = public)
 
 		% CONSTRUCTOR =====================================================
-		function obj = ModelSeperate(toolboxPath, sampler, data, saveFolder)
+		function obj = ModelSeparate(toolboxPath, sampler, data, saveFolder)
 			% Because this class is a subclass of "ModelBaseClass" then we use
 			% this next line to create an instance
 			obj = obj@ModelBaseClass(toolboxPath, sampler, data, saveFolder);
 
 			switch sampler
 				case{'JAGS'}
-					obj.sampler = JAGSSampler([toolboxPath '/jagsModels/seperateME.txt'])
+					obj.sampler = JAGSSampler([toolboxPath '/jagsModels/separateME.txt'])
 					[~,obj.modelType,~] = fileparts(obj.sampler.fileName);
 				case{'STAN'}
 					error('NOT IMPLEMENTED YET')
@@ -29,40 +29,40 @@
 			%% Create variables
 			% -------------------------------------------------------------------
 			% Participant-level -------------------------------------------------
-			m = Variable('m','m', [], true)
+			m = Variable('m','m', [], true);
 			m.seed.func = @() normrnd(-0.243,10);
 			m.seed.single = false;
 
-			mprior = Variable('mprior','mprior', [], true)
+			mprior = Variable('mprior','mprior', [], true);
 			mprior.seed.func = @() normrnd(-0.243,10);
 			mprior.seed.single = true;
 
-			c = Variable('c','c', [], true)
+			c = Variable('c','c', [], true);
 			c.seed.func = @() normrnd(0,10);
 			c.seed.single = false;
 
-			cprior = Variable('cprior','cprior', [], true)
+			cprior = Variable('cprior','cprior', [], true);
 			cprior.seed.func = @() normrnd(0,10);
 			cprior.seed.single = true;
 
-			epsilon = Variable('epsilon','\epsilon', [0 0.5], true)
+			epsilon = Variable('epsilon','\epsilon', [0 0.5], true);
 			epsilon.seed.func = @() rand/2;
 			epsilon.seed.single = false;
 
-			epsilonprior = Variable('epsilonprior','\epsilon prior', [0 0.5], true)
+			epsilonprior = Variable('epsilonprior','\epsilon prior', [0 0.5], true);
 			epsilonprior.seed.func = @() rand/2;
 			epsilonprior.seed.single = true;
 
-			alpha = Variable('alpha','\alpha', 'positive', true)
+			alpha = Variable('alpha','\alpha', 'positive', true);
 			alpha.seed.func = @() abs(normrnd(0,10));
 			alpha.seed.single = false;
 
-			alphaprior = Variable('alphaprior','\alpha prior', 'positive', true)
+			alphaprior = Variable('alphaprior','\alpha prior', 'positive', true);
 			alphaprior.seed.func = @() abs(normrnd(0,10));
 			alphaprior.seed.single = true;
 
 			% posterior predictive ----------------------------------------------
-			Rpostpred = Variable('Rpostpred','Rpostpred', [0 1], true)
+			Rpostpred = Variable('Rpostpred','Rpostpred', [0 1], true);
 			Rpostpred.plotMCMCchainFlag = false;
 
 			% define which to analyse (univariate analysis)

demo/SCRIPT.m

@@ -122,23 +122,25 @@
 % order to have more reliable MCMC chain convergence.
 % This new model is called ModelHierarchicalUpdated
 
-saveFolder = 'hierarchical_updated_priors';
-uModel = ModelHierarchicalUpdated(toolboxPath, 'JAGS', myData, saveFolder);
-uModel.sampler.setMCMCtotalSamples(10^5);
-uModel.conductInference();
-uModel.posteriorPredictive();
-uModel.exportParameterEstimates();
-uModel.plot()
-uModel.plotMCMCchains()
-
-
-%% CODE FOR PARTICIPANT-LEVEL ONLY INFERENCE IS EXPERIMENTAL... NOT RECOMMENDED
-% % PARTICIPANT-LEVEL ONLY INFERENCES
+% saveFolder = 'hierarchical_updated_priors';
+% uModel = ModelHierarchicalUpdated(toolboxPath, 'JAGS', myData, saveFolder);
+% uModel.sampler.setMCMCtotalSamples(10^5);
+% uModel.conductInference();
+% uModel.posteriorPredictive();
+% uModel.exportParameterEstimates();
+% uModel.plot()
+% uModel.plotMCMCchains()
+
+
+%% CODE FOR PARTICIPANT-LEVEL ONLY
+% THIS IS CURRENTLY EXPERIMENTAL. IT WORKS FINE WITH THE DEMO DATASET, BUT
+% I HAVE BEEN HAVING MCMC CHAIN CONVERGENCE ISSUES WITH OTHER DATASETS.
+
 % % If you want to avoid group-level hierarchical inference, then you can use
 % % a different model class. Code below shows an example
 % 
 % saveFolder = 'nonHierarchical';
-% sModel = ModelSeperate(toolboxPath, 'JAGS', myData, saveFolder);
+% sModel = ModelSeparate(toolboxPath, 'JAGS', myData, saveFolder);
 % sModel.sampler.setMCMCtotalSamples(10^5);
 % sModel.conductInference();
 % sModel.posteriorPredictive();