You can post-process and analyze an existing ParaMonte simulation in MATLAB irrespective of when and how this simulation was generated by any of the available ParaMonte library interfaces (e.g., C/C++, Fortran, MATLAB, Python, …). This is one of the many unique features of the ParaMonte library.
The ParaMonte MATLAB library is particularly suitable for post-processing the output of large-scale high-performance ParaMonte simulations performed in C/C++/Fortran.
Post-processing an existing ParaDRAM simulation
To analyze an existing ParaDRAM simulation, all you need to do is to specify the full-filename or minimally, the prefix by which the simulation output filenames have been created. For example, suppose we want to reanalyze this output mvn_serial_process_1_sample.txt sample file from an existing simulation. All that we need to do to read this file, is to pass the local path to this file on your system to an instance of the ParaDRAM class of the ParaMonte MATLAB library.
Suppose we have stored this file in "./temp/mvn_serial_process_1_sample.txt" locally on the system, with respect to the current working directory of the MATLAB session. Then,
addpath(genpath("./"),"-begin"); % add the ParaMonte library directories to MATLAB's list of search paths.
pm = paramonte(); % instantiate an object of class paramonte.
pmpd = pm.ParaDRAM(); % instantiate an object of class ParaDRAM.
pmpd.readSample("./temp/mvn_serial_process_1_sample.txt"); % read the downloaded sample file and store it in the newly-created pmpd.sampleList property.
head(pmpd.sampleList{1}.df) % this newly-created pmpd.sampleList MATLAB-cell property contains the contents of the sample file and other relevant tools.
readSample(), readChain(), readMarkovChain(), readProgress(), readRestart(), readReport() of the ParaDRAM sampler, can also take a URL as the input path to the file. In such case, a temporary copy of the file will be downloaded from the web and removed once its contents are parsed by the parser method.Done. The rest of the postprocessing of the simulation data, such as statistical analysis or plotting, is identical to the MATLAB examples provided on this page.
,), we can specify the delimiter by passing it explicitly,
addpath(genpath("./"),"-begin"); % add the ParaMonte library directories to MATLAB's list of search paths.
pm = paramonte(); % instantiate an object of class paramonte.
pmpd = pm.ParaDRAM(); % instantiate an object of class ParaDRAM.
ppmpd.readSample("./temp/mvn_serial_process_1_sample.txt", ","); % read the downloaded sample file by explicitly passing the delimiter used in the sample file.Alternative method:
Alternatively, we could also specify both the filename and delimiter as attributes of the pmpd object and avoid passing them directly to the pmpd.readSample() method,
addpath(genpath("./"),"-begin"); % add the ParaMonte library directories to MATLAB's list of search paths.
pm = paramonte(); % instantiate an object of class paramonte.
pmpd = pm.ParaDRAM(); % instantiate an object of class ParaDRAM.
pmpd.spec.outputFileName = "./temp/mvn_serial_process_1_sample.txt";
pmpd.spec.outputDelimiter = ",";
pmpd.readSample(); % read the downloaded sample file and store it in the newly-created pmpd.sampleList property.
head(pmpd.sampleList{1}.df) % this newly-created MATLAB-cell property now contains the contents of the sample file and other relevant tools.