When initiating model parameter estimation, there is only one required input, which is the a data. So you can get up and running very quickly with a simple call like this…
myData = Data(datapath, 'files', allFilesInFolder(datapath, 'txt'));
model = ModelHierarchicalME(myData);
The function being called constructs a Model object and then proceeds to do parameter estimation. The toolbox provides a number of models to use, and it is important to pick the one appropriate for your data and research question. See the Models page for more information.
Data objectThe first input has to be a Data object. This can be constructed as follows
Data(datapath, 'files', allFilesInFolder(datapath, 'txt'))
The first argument must be a string, specifying the location of data files. We also need one key/value pair:
| Key | Values |
|---|---|
'files' |
must be a cell array of strings |
the input can either be a manually specified cell array of strings, or the output of a utility function that yields a cell array of strings. Here, I use a helper function allFilesInFolder() which outputs a cell array of strings, namely all of the .txt files in the specified datapath. You can feel free to write your own utility function, as long as the input is a cell array of strings, it should work.
The rest of the arguments are optional key/value pairs
| Key | Values |
|---|---|
'timeUnits' |
'minutes' or 'hours' or 'days' [default] |
'pointEstimateType' |
'mean' or 'median' or 'mode' [default] |
'shouldPlot' |
'yes' or 'no' [default] |
'shouldExportPlots' |
'true' [default] or 'false' |
'exportFormats' |
a cell array of output formats, e.g. {'png', 'pdf'} (default is {'png'} only) |
'shouldExportPlots' |
true or false |
'savePath' |
a string defining the folder location to save exports in |
'mcmcParams' |
a structure (see below) |
You can override the default MCMC sampling parameters by providing a structure in the form of:
struct('nsamples', 10^5,...
'nchains', 4,...
'nburnin', 10^3)
Please see run_me.m for a working example. But here is a quick example of how to conduct the parameter estimation
model = ModelHierarchicalME(Data(datapath, 'files', allFilesInFolder(datapath, 'txt')),...
'timeUnits', 'days',...
'savePath', fullfile(pwd,'output','my_analysis'),...
'pointEstimateType', 'mode',...
'shouldPlot', 'no',...
'shouldExportPlots', false,...
'exportFormats', {'png'},...
'mcmcParams', struct('nsamples', 100000, 'nchains', 4, 'nburnin', 1000));