Create weighted RooDataset with RDataframe

Roofit and RooStats
,
1

Hi ROOTers,

I have encountered a problem while creating a RooDataSet using the RDataFrame interface. I followed the tutorial and wrote the following code:

RooRealVar Ks_M("Ks_M", "Ks Mass", fit_min, fit_max);
RooRealVar sb_weight("sb_weight", "Sideband weight", 0.0);

auto filtered_df = df
    .Filter(All_Cut.GetTitle())
    .Define("sb_weight", []() { return -0.6; });

auto data_weighted = filtered_df.Book<double>(
    RooDataSetHelper{
        "weighted_dataset",
        "weighted_Dataset",
        RooArgSet(Ks_M),
    },
    {"Ks_M", "sb_weight"}
);

However, I receive the following error:

terminate called after throwing an instance of 'std::runtime_error'
  what():  1 column name is required but 2 were provided: "Ks_M", "sb_weight".

Please let me know if you have any idea.

Best regards,
bshi

ROOT version: 6.32.02

2

Hi @bshi,

Thanks for your question. I’m adding @vpadulan in the loop.

Cheers,
Dev

3

Did you try removing “sb_weight” from the last line?

or alternatively maybe adding sb_weight to rooargset (and adding second double) in Book<>

4

Hi @ferhue,

Thank you for your help. I implemented the changes as you recommended, and here is the updated code:

auto data_weighted = filtered_df.Book<double, double>(
    RooDataSetHelper{
                    "weighted dataset",  
                    "weighted Dataset",  
                    RooArgSet(Ks_M, sb_weight)},  
    {"Ks_M"}  
);

However, I encountered the following error:

terminate called after throwing an instance of 'std::runtime_error'  
  what():  2 column names are required but 1 was provided: "Ks_M".

To resolve this, I modified the last line to include both columns:

{"Ks_M", "sb_weight"}

The code now runs without errors, but it doesn’t produce the intended result. Specifically, when I calculate the sum of weights using the following code:

double sum_weight = 0.0;  
for (int i = 0; i < data_weighted->numEntries(); ++i) {  
    sum_weight += data_weighted->weight();  
}  
std::cout << "Sum of weights = " << sum_weight << std::endl;

The output is:

Sum of weights = 55

This indicates that the weighting is not being applied correctly, as the original number of events is 55, and I included negative weights in the dataset.

Additionally, I performed a test using the following code:

auto data_weighted = filtered_df.Book<double, double>(  
    RooDataSetHelper{  
        "weighted dataset",  
        "weighted Dataset",  
        RooArgSet(Ks_M),  
        RooFit::WeightVar("sb_weight")  
    },  
    {"Ks_M", "sb_weight"}  
);

However, the error shows:

terminate called after throwing an instance of 'std::invalid_argument'  
  what():  RooDataSet can hold 1 variables per event, but RDataFrame passed 2 columns.

I look forward to your more idea on this issue.

Thank you!

Best regards,
bshi