Serialization

Once you have a PointProcessResult (output of RegressionPipeline::fullRegression() you can use the following functions to serialize both the regression parameters at each time step (ALERT: The .csv file generated tends to be quite large, e.g. 1 GB for 3 hours of data processed with a delta = 0.002) and the Taus (integral of hazard-rate for each observed event) in order to assess goodness of fit.

// ppRes is a PointProcessResult obtained by running RegressionPipeline::fullRegression()
ppResData2csv(ppRes, std::string("myData.csv"));
ppResTaus2csv(ppRes, std::string("myTaus.csv"));

The header of these files (supposing we processed the data with an AR_ORDER = 2) will be: