Logo ROOT  
Reference Guide
rf408_RDataFrameToRooFit.C File Reference

Detailed Description

View in nbviewer Open in SWAN Fill RooDataSet/RooDataHist in RDataFrame.

This tutorial shows how to fill RooFit data classes directly from RDataFrame. Using two small helpers, we tell RDataFrame where the data has to go.

#include <TROOT.h>
#include <TRandom.h>
#include <initializer_list>
{
// Set up
// ------------------------
// We enable implicit parallelism, so RDataFrame runs in parallel.
// We create an RDataFrame with two columns filled with 2 million random numbers.
ROOT::RDataFrame d(2000000);
auto dd = d.Define("x", [](){ return gRandom->Uniform(-5., 5.); })
.Define("y", [](){ return gRandom->Gaus(1., 3.); });
// We create RooFit variables that will represent the dataset.
RooRealVar x("x", "x", -5., 5.);
RooRealVar y("y", "y", -50., 50.);
x.setBins(10);
y.setBins(20);
// Booking the creation of RooDataSet / RooDataHist in RDataFrame
// ----------------------------------------------------------------
// Method 1:
// We directly book the RooDataSetMaker action.
// We need to pass
// - the RDataFrame column types as template parameters
// - the constructor arguments for RooDataSet (they follow the same syntax as the usual RooDataSet constructors)
// - the column names that RDataFrame should fill into the dataset
//
// NOTE: RDataFrame columns are matched to RooFit variables by position, *not by name*!
auto rooDataSet = dd.Book<double, double>(
RooDataSetHelper("dataset", // Name
"Title of dataset", // Title
RooArgSet(x, y) // Variables in this dataset
),
{"x", "y"} // Column names in RDataFrame.
);
// Method 2:
// We first declare the RooDataHistMaker
RooDataHistHelper rdhMaker{"datahist", // Name
"Title of data hist", // Title
RooArgSet(x, y) // Variables in this dataset
};
// Then, we move it into the RDataFrame action:
auto rooDataHist = dd.Book<double, double>(std::move(rdhMaker), {"x", "y"});
// Run it and inspect the results
// -------------------------------
// Let's inspect the dataset / datahist.
// Note that the first time we touch one of those objects, the RDataFrame event loop will run.
for (const RooAbsData* data : std::initializer_list<const RooAbsData*>{rooDataSet.GetPtr(), rooDataHist.GetPtr()} ) {
std::cout << std::endl;
data->Print();
for (int i=0; i < data->numEntries() && i < 20; ++i) {
std::cout << "(";
for (auto var : *data->get(i)) {
std::cout << std::setprecision(3) << std::right << std::fixed << std::setw(8) << static_cast<const RooAbsReal*>(var)->getVal() << ", ";
}
std::cout << ")\tweight=" << std::setw(10) << data->weight() << std::endl;
}
std::cout << "mean(x) = " << data->mean(x) << "\tsigma(x) = " << std::sqrt(data->moment(x, 2.))
<< "\n" << "mean(y) = " << data->mean(y) << "\tsigma(y) = " << std::sqrt(data->moment(y, 2.)) << std::endl;
}
}
int main() {
return 0;
}
double
Definition: Converters.cxx:939
#define d(i)
Definition: RSha256.hxx:102
RooAbsDataHelper< RooDataSet > RooDataSetHelper
Helper for creating a RooDataSet inside RDataFrame.
double sqrt(double)
R__EXTERN TRandom * gRandom
Definition: TRandom.h:62
int main(int argc, char *argv[])
Definition: cef_main.cxx:54
ROOT's RDataFrame offers a high level interface for analyses of data stored in TTree,...
Definition: RDataFrame.hxx:40
This is a helper for an RDataFrame action, which fills RooFit data classes.
RooAbsData is the common abstract base class for binned and unbinned datasets.
Definition: RooAbsData.h:79
RooArgSet is a container object that can hold multiple RooAbsArg objects.
Definition: RooArgSet.h:35
RooRealVar represents a variable that can be changed from the outside.
Definition: RooRealVar.h:39
virtual Double_t Gaus(Double_t mean=0, Double_t sigma=1)
Samples a random number from the standard Normal (Gaussian) Distribution with the given mean and sigm...
Definition: TRandom.cxx:274
virtual Double_t Uniform(Double_t x1=1)
Returns a uniform deviate on the interval (0, x1).
Definition: TRandom.cxx:672
Double_t y[n]
Definition: legend1.C:17
Double_t x[n]
Definition: legend1.C:17
void EnableImplicitMT(UInt_t numthreads=0)
Enable ROOT's implicit multi-threading for all objects and methods that provide an internal paralleli...
Definition: TROOT.cxx:527
␛[1mRooFit v3.60 -- Developed by Wouter Verkerke and David Kirkby␛[0m
Copyright (C) 2000-2013 NIKHEF, University of California & Stanford University
All rights reserved, please read http://roofit.sourceforge.net/license.txt
RooDataSet::dataset[x,y] = 2000000 entries
( 4.997, -0.304, ) weight= 1.000
( 4.472, 0.910, ) weight= 1.000
( 4.575, 0.830, ) weight= 1.000
( 0.400, 0.776, ) weight= 1.000
( 2.599, -0.232, ) weight= 1.000
( -1.844, 1.575, ) weight= 1.000
( 0.197, 0.853, ) weight= 1.000
( -1.077, -0.721, ) weight= 1.000
( -4.697, -3.165, ) weight= 1.000
( 4.437, -1.208, ) weight= 1.000
( 3.983, -0.146, ) weight= 1.000
( -0.014, -1.447, ) weight= 1.000
( -3.177, -2.704, ) weight= 1.000
( -4.371, -0.363, ) weight= 1.000
( 2.254, -0.499, ) weight= 1.000
( 2.139, 6.533, ) weight= 1.000
( 1.993, 6.991, ) weight= 1.000
( -3.708, 7.781, ) weight= 1.000
( -4.168, 1.284, ) weight= 1.000
( -4.177, 4.650, ) weight= 1.000
mean(x) = 0.001 sigma(x) = 2.888
mean(y) = 1.001 sigma(y) = 2.999
RooDataHist::datahist[x,y] = 200 bins (2000000.000 weights)
( -4.500, -47.500, ) weight= 0.000
( -4.500, -42.500, ) weight= 0.000
( -4.500, -37.500, ) weight= 0.000
( -4.500, -32.500, ) weight= 0.000
( -4.500, -27.500, ) weight= 0.000
( -4.500, -22.500, ) weight= 0.000
( -4.500, -17.500, ) weight= 0.000
( -4.500, -12.500, ) weight= 28.000
( -4.500, -7.500, ) weight= 4252.000
( -4.500, -2.500, ) weight= 70132.000
( -4.500, 2.500, ) weight=106989.000
( -4.500, 7.500, ) weight= 18635.000
( -4.500, 12.500, ) weight= 239.000
( -4.500, 17.500, ) weight= 0.000
( -4.500, 22.500, ) weight= 0.000
( -4.500, 27.500, ) weight= 0.000
( -4.500, 32.500, ) weight= 0.000
( -4.500, 37.500, ) weight= 0.000
( -4.500, 42.500, ) weight= 0.000
( -4.500, 47.500, ) weight= 0.000
mean(x) = 0.001 sigma(x) = 2.873
mean(y) = 1.003 sigma(y) = 3.330
Date
Mar 2021
Author
Stephan Hageboeck (CERN)

Definition in file rf408_RDataFrameToRooFit.C.