RooMCStudy error

[Vikas_Raj]

Dear experts,

I am trying to generate pure toy datasets using RooMCStudy::generateAndFit() for a combined dataset and a simultaneous PDF, which is composed of individual PDFs defined in 7 different bins of a flavor tagging parameter r.

I am using the following code:

reso_ws.factory("sample_4d_pure[r0,r1,r2,r3,r4,r5,r6]")

reso_ws.factory("SIMUL::simPdf_TD(sample_4d_pure, "
           "r0=pdf_full_td0, "
           "r1=pdf_full_td1, "
           "r2=pdf_full_td2, "
           "r3=pdf_full_td3, "
           "r4=pdf_full_td4, "
           "r5=pdf_full_td5, "
           "r6=pdf_full_td6)"
           )

print(reso_ws.obj("simPdf_TD"))

dataset_td = R.RooDataSet("dataset_td", "combined dataset", data_set_td_0.get())
for i in range(7):
    ds = globals()[f"data_set_td_{i}"]
    dataset_td.append(ds)

dataset_td.Print('v')


mcstudy = R.RooMCStudy(
    reso_ws.pdf("simPdf_TD"),
    obs_td,
    R.RooFit.ProtoData(dataset_td),
    # Binned=True,
    Extended=True,
    # Silence=True,
    FitOptions=dict(Save=True, PrintEvalErrors=1, NumCPU=10)
)


mcstudy.generate(5)

Here is the error log:
roomcstudy_errot.txt (7.3 KB)

Could you please help me correct or improve it?

Thanks!

[Danilo]

Hi,

Thanks for the post. The code does not seem to be enough to reproduce the error, but the messages seem to be clear enough - Did you try to remove the duplicates as a start?

Cheers,
Danilo

[Vikas_Raj]

Hello!
Thanks for the reply!

@Danilo

• Did you try to remove the duplicates as a start?

Regarding this, I am not sure how to remove duplicates here as I am using a simultaneous PDF of seven PDFs in different r bins, and they all have same variables as observables

obs_td = {
    reso_ws.var("deltat"), 
    reso_ws.var("deltaterr"), 
    reso_ws.var("r"),
    reso_ws.cat("tagflav"),
    reso_ws.var("mod_mbc"), 
    reso_ws.var("deltae"),
    reso_ws.var("csobdtprime")
}

data_set_td_0 = R.RooDataSet.from_pandas(data_td0, obs_td)
data_set_td_1 = R.RooDataSet.from_pandas(data_td1, obs_td)
data_set_td_2 = R.RooDataSet.from_pandas(data_td2, obs_td)
data_set_td_3 = R.RooDataSet.from_pandas(data_td3, obs_td)
data_set_td_4 = R.RooDataSet.from_pandas(data_td4, obs_td)
data_set_td_5 = R.RooDataSet.from_pandas(data_td5, obs_td)
data_set_td_6 = R.RooDataSet.from_pandas(data_td6, obs_td)

Dear @Danilo
Please find the minimal reproducer attached below.
toys_mini.py (1.3 KB)