Dear experts,
Greetings,
Currently, I try to produce Toy MC from binned data. if there is only one category, I can do it by
// RooWorkspace* w
ModelConfig* mc = (ModelConfig*)w->obj("ModelConfig"); // Get model manually
RooArgSet* obs = (RooArgSet*)mc->GetObservables();
RooRealVar* x_val_MXs1 = (RooRealVar*)obs->find("obs_x_channel_MXs1");
RooDataSet* data = (RooDataSet*)w->data("asimovData");
TH1D* HistData = (TH1D*)data->createHistogram("HistData", *x_val_MXs1, Binning(RarityBins, BinMIN, BinMAX));
RooDataHist DataHist("DataHist", "DataHist", *x_val_MXs1, HistData);
RooHistPdf dataPDF("dataPDF", "dataPDF", RooArgSet(*x_val_MXs1, model->indexCat()), DataHist);
RooDataSet* genData = dataPDF.generate(RooArgSet(*x_val_MXs1, model->indexCat()), 100, false, true, "", false, true);
The above code try to do
RooDataSet (asimovData) → TH1D → RooDataHist → RooHistPdf → RooDataSet
However, if there are more than one RooCategory, it is not simple:
The ranges and number of bins are taken from the input histogram and must be the same in all histograms
https://root.cern.ch/doc/master/classRooDataHist.html#a60ffc4da687de2eee9ac1696725fc175
However, for each category, the bin and range can be different.
// RooWorkspace* w
ModelConfig* mc = (ModelConfig*)w->obj("ModelConfig"); // Get model manually
RooArgSet* obs = (RooArgSet*)mc->GetObservables();
RooRealVar* x_val_MXs1 = (RooRealVar*)obs->find("obs_x_channel_MXs1");
RooRealVar* x_val_MXs2 = (RooRealVar*)obs->find("obs_x_channel_MXs2");
RooRealVar* x_val_MXs3 = (RooRealVar*)obs->find("obs_x_channel_MXs3");
// ...?
So, what is the best way to produce MC from data (here, from asimovData)
Thank you