# Unexpected result with RooFFTConvPDF after ROOT 6.32

**URL:** <https://root-forum.cern.ch/t/unexpected-result-with-roofftconvpdf-after-root-6-32/64574>\
**Category:** Roofit and RooStats\
**Tags:** root\
**Created:** [January 7, 2026, 3:48pm UTC](https://root-forum.cern.ch/t/unexpected-result-with-roofftconvpdf-after-root-6-32/64574 "2026-01-07T15:48:11Z")\
**Posts on this page:** 7\
**Page:** 1

<div class="post-metadata">

**Author:** ![kyu](https://root-forum.cern.ch/letter_avatar_proxy/v4/letter/k/b3f665/32.png) [@kyu](https://root-forum.cern.ch/u/kyu)\
**Post date:** [January 7, 2026, 3:48pm UTC](https://root-forum.cern.ch/t/unexpected-result-with-roofftconvpdf-after-root-6-32/64574/1 "2026-01-07T15:48:11Z")

</div>

Dear experts,

I am using the RooFFTConvPdf to describe the detector resolution.  
I found that the fit could easily converge with the root before v6.32.00.  
But diverge for the v6.32.00.  
Given the log

```auto
RooAbsMinimizerFcn: Minimized function has error status.
Returning maximum FCN so far (-inf) to force MIGRAD to back out of this region. Error log follows.
Parameter values: c0=-0.491664	mean=5800	n_CombBkg=358	n_argus=305.5	nsig=100	sigma=8.09227
RooAddPdf::model[nsig * sigpdf_over_sigpdf_Int[mass] + n_CombBkg * CombBkg_over_CombBkg_Int[mass] + n_argus * Part_Rec_Bkg_over_Part_Rec_Bkg_Int[mass] ]
     getLogVal() top-level p.d.f evaluates to NaN @ !refCoefNorm=(mass = 5800), !pdfs=(sigpdf_over_sigpdf_Int[mass] = 0.0492992,CombBkg_over_CombBkg_Int[mass] = 0,Part_Rec_Bkg_over_Part_Rec_Bkg_Int[mass] = 7.66857e-19), !coefficients=(nsig = 100 +/- 0,n_CombBkg = 358 +/- 0,n_argus = 305.5 +/- 0)
     getLogVal() top-level p.d.f evaluates to NaN @ !refCoefNorm=(mass = 5800), !pdfs=(sigpdf_over_sigpdf_Int[mass] = 0.0492992,CombBkg_over_CombBkg_Int[mass] = 0,Part_Rec_Bkg_over_Part_Rec_Bkg_Int[mass] = 7.66857e-19), !coefficients=(nsig = 100 +/- 0,n_CombBkg = 358 +/- 0,n_argus = 305.5 +/- 0)
     getLogVal() top-level p.d.f evaluates to NaN @ !refCoefNorm=(mass = 5800), !pdfs=(sigpdf_over_sigpdf_Int[mass] = 0.0492992,CombBkg_over_CombBkg_Int[mass] = 0,Part_Rec_Bkg_over_Part_Rec_Bkg_Int[mass] = 7.66857e-19), !coefficients=(nsig = 100 +/- 0,n_CombBkg = 358 +/- 0,n_argus = 305.5 +/- 0)
     getLogVal() top-level p.d.f evaluates to NaN @ !refCoefNorm=(mass = 5800), !pdfs=(sigpdf_over_sigpdf_Int[mass] = 0.0492992,CombBkg_over_CombBkg_Int[mass] = 0,Part_Rec_Bkg_over_Part_Rec_Bkg_Int[mass] = 7.66857e-19), !coefficients=(nsig = 100 +/- 0,n_CombBkg = 358 +/- 0,n_argus = 305.5 +/- 0)
     getLogVal() top-level p.d.f evaluates to NaN @ !refCoefNorm=(mass = 5800), !pdfs=(sigpdf_over_sigpdf_Int[mass] = 0.0492992,CombBkg_over_CombBkg_Int[mass] = 0,Part_Rec_Bkg_over_Part_Rec_Bkg_Int[mass] = 7.66857e-19), !coefficients=(nsig = 100 +/- 0,n_CombBkg = 358 +/- 0,n_argus = 305.5 +/- 0)
     getLogVal() top-level p.d.f evaluates to NaN @ !refCoefNorm=(mass = 5800), !pdfs=(sigpdf_over_sigpdf_Int[mass] = 0.0492992,CombBkg_over_CombBkg_Int[mass] = 0,Part_Rec_Bkg_over_Part_Rec_Bkg_Int[mass] = 7.66857e-19), !coefficients=(nsig = 100 +/- 0,n_CombBkg = 358 +/- 0,n_argus = 305.5 +/- 0)
     getLogVal() top-level p.d.f evaluates to NaN @ !refCoefNorm=(mass = 5800), !pdfs=(sigpdf_over_sigpdf_Int[mass] = 0.0492992,CombBkg_over_CombBkg_Int[mass] = 0,Part_Rec_Bkg_over_Part_Rec_Bkg_Int[mass] = 7.66857e-19), !coefficients=(nsig = 100 +/- 0,n_CombBkg = 358 +/- 0,n_argus = 305.5 +/- 0)
     getLogVal() top-level p.d.f evaluates to NaN @ !refCoefNorm=(mass = 5800), !pdfs=(sigpdf_over_sigpdf_Int[mass] = 0.0492992,CombBkg_over_CombBkg_Int[mass] = 0,Part_Rec_Bkg_over_Part_Rec_Bkg_Int[mass] = 7.66857e-19), !coefficients=(nsig = 100 +/- 0,n_CombBkg = 358 +/- 0,n_argus = 305.5 +/- 0)
     getLogVal() top-level p.d.f evaluates to NaN @ !refCoefNorm=(mass = 5800), !pdfs=(sigpdf_over_sigpdf_Int[mass] = 0.0492992,CombBkg_over_CombBkg_Int[mass] = 0,Part_Rec_Bkg_over_Part_Rec_Bkg_Int[mass] = 7.66857e-19), !coefficients=(nsig = 100 +/- 0,n_CombBkg = 358 +/- 0,n_argus = 305.5 +/- 0)
     getLogVal() top-level p.d.f evaluates to NaN @ !refCoefNorm=(mass = 5800), !pdfs=(sigpdf_over_sigpdf_Int[mass] = 0.0492992,CombBkg_over_CombBkg_Int[mass] = 0,Part_Rec_Bkg_over_Part_Rec_Bkg_Int[mass] = 7.66857e-19), !coefficients=(nsig = 100 +/- 0,n_CombBkg = 358 +/- 0,n_argus = 305.5 +/- 0)
     getLogVal() top-level p.d.f evaluates to NaN @ !refCoefNorm=(mass = 5800), !pdfs=(sigpdf_over_sigpdf_Int[mass] = 0.0492992,CombBkg_over_CombBkg_Int[mass] = 0,Part_Rec_Bkg_over_Part_Rec_Bkg_Int[mass] = 7.66857e-19), !coefficients=(nsig = 100 +/- 0,n_CombBkg = 358 +/- 0,n_argus = 305.5 +/- 0)
     getLogVal() top-level p.d.f evaluates to NaN @ !refCoefNorm=(mass = 5800), !pdfs=(sigpdf_over_sigpdf_Int[mass] = 0.0492992,CombBkg_over_CombBkg_Int[mass] = 0,Part_Rec_Bkg_over_Part_Rec_Bkg_Int[mass] = 7.66857e-19), !coefficients=(nsig = 100 +/- 0,n_CombBkg = 358 +/- 0,n_argus = 305.5 +/- 0)
    ... (remaining 18 messages suppressed)

Warning in <Minuit2>: MnHesse 2nd derivative zero for parameter c0 ; MnHesse fails and will return diagonal matrix
Info in <Minuit2>: MnHesse Done after 10.2978 ms
Warning in <Minuit2>: Minuit2Minimizer::Hesse Hesse failed - matrix is full but not positive defined
Warning in <Minuit2>: Minuit2Minimizer::Hesse 3
[#0] ERROR:Minimization -- RooMinimizer::calculateHessErrors() Error when calculating Hessian
[#0] ERROR:Minimization -- RooMinimizer: all function calls during minimization gave invalid NLL values!
[#1] INFO:Minimization -- RooAbsMinimizerFcn::setOptimizeConst: deactivating const optimization

  RooFitResult: minimized FCN value: 1.98025e-13, estimated distance to minimum: 6.06841e-16
                covariance matrix quality: Not calculated at all
                Status : MINIMIZE=-1 HESSE=302 

```

When I am looking at the [ROOT Version 6.32 Release Notes](https://root.cern/doc/v632/release-notes.html) , it said

> I f you experience unexpected problems related to the likelihood evaluation, you can revert back to the old backend by passing `RooFit::EvalBackend("legacy")` to `RooAbsPdf::fitTo()` or `RooAbsPdf::createNLL()`.

By adding the argument `RooFit::EvalBackend("legacy")`, the proble was fixed.  
And the fit result agree with different root version.  
Is that under the expectation?

Here is the working example, I think it could be reproduced.  
[MWE\_argus.C](https://root-forum.cern.ch/uploads/short-url/ezDRvZWLuZ8U07jCiBKUhu949qa.C) (4.1 KB)

- platform: AlmaLinux9
- root version: 6.30.08 and 6.32.00 from LCG in cvmfs

Many thanks.

---

<div class="post-metadata">

**Author:** ![Danilo](https://root-forum.cern.ch/user_avatar/root-forum.cern.ch/danilo/32/34528_2.png) [@Danilo](https://root-forum.cern.ch/u/Danilo)\
**Post date:** [January 7, 2026, 7:58pm UTC](https://root-forum.cern.ch/t/unexpected-result-with-roofftconvpdf-after-root-6-32/64574/2 "2026-01-07T19:58:16Z")

</div>

Hi Kyu,

Thanks for sharing with the community this example. We are looking into it and will come back to you (@jonas)

Cheers,  
D

---

<div class="post-metadata">

**Author:** ![kyu](https://root-forum.cern.ch/letter_avatar_proxy/v4/letter/k/b3f665/32.png) [@kyu](https://root-forum.cern.ch/u/kyu)\
**Post date:** [January 8, 2026, 3:21am UTC](https://root-forum.cern.ch/t/unexpected-result-with-roofftconvpdf-after-root-6-32/64574/3 "2026-01-08T03:21:53Z")

</div>

Sorry that I made a mistake in the attachment before.  
The `argus_m0` should be `5800 - 135`, not `5800, 135`. The attached code is updated now.

The problem seems minor in this work example, but much terrible when I am doing for the real data. I can not even got a resonable fit result.  
I am sorrry I can not put the plot and code for the real data case.  
But it have the same structure on the PDF construction and it shows similar log to this working example.

Best regards,  
Kunpeng

---

<div class="post-metadata">

**Author:** ![Danilo](https://root-forum.cern.ch/user_avatar/root-forum.cern.ch/danilo/32/34528_2.png) [@Danilo](https://root-forum.cern.ch/u/Danilo)\
**Post date:** [January 8, 2026, 6:08am UTC](https://root-forum.cern.ch/t/unexpected-result-with-roofftconvpdf-after-root-6-32/64574/4 "2026-01-08T06:08:13Z")

</div>

Thanks Kyu, your constraints are perfectly reasonable, no worries. A reproducer with fake data should be enough.

Cheers,  
D

---

<div class="post-metadata">

**Author:** ![system](https://root-forum.cern.ch/uploads/default/original/3X/e/f/efe4f24398fd8ffeb76bccb6ecc46d9d9c003376.png) [@system](https://root-forum.cern.ch/u/system)\
**Post date:** [January 22, 2026, 6:08am UTC](https://root-forum.cern.ch/t/unexpected-result-with-roofftconvpdf-after-root-6-32/64574/5 "2026-01-22T06:08:17Z")

</div>

Closed on January 22, 2026, 6:08am UTC

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.

---

<div class="post-metadata">

**Author:** ![jonas](https://root-forum.cern.ch/user_avatar/root-forum.cern.ch/jonas/32/24756_2.png) [@jonas](https://root-forum.cern.ch/u/jonas)\
**Post date:** [February 2, 2026, 2:16pm UTC](https://root-forum.cern.ch/t/unexpected-result-with-roofftconvpdf-after-root-6-32/64574/6 "2026-02-02T14:16:51Z")

</div>

Opened on February 2, 2026, 2:16pm UTC



---

<div class="post-metadata">

**Author:** ![jonas](https://root-forum.cern.ch/user_avatar/root-forum.cern.ch/jonas/32/24756_2.png) [@jonas](https://root-forum.cern.ch/u/jonas)\
**Post date:** [February 4, 2026, 2:45pm UTC](https://root-forum.cern.ch/t/unexpected-result-with-roofftconvpdf-after-root-6-32/64574/7 "2026-02-04T14:45:11Z")

</div>

Thanks for the report! This is indeed unexpected, and related to the new vectorized likelihood evaluation backend in ROOT 6.32.

I have opened a GitHub issue about this and hope to fix it for the next ROOT patch releases:

> <https://github.com/root-project/root/issues/21159>
>
> Another corner case where the new RooFit evaluation backend fails to give the ri…ght results was reported on the ROOT forum:
> 
> https://root-forum.cern.ch/t/unexpected-result-with-roofftconvpdf-after-root-6-32/64574/4
> 
> Reproducer:
> \`\`\`txt
> using namespace RooFit;
> 
> RooRealVar obs{"mass", "", 5500, 6100};
> RooRealVar mean{"mean", "", 5800.0, 5795, 5805};
> RooRealVar sigma{"sigma", "", 8.1, 1, 30};
> 
> RooGaussian sigpdf{"sigpdf", "", obs, mean, sigma};
> 
> RooRealVar argus\_par{"argus\_par", "", -55.0};
> RooRealVar argus\_m0{"argus\_m0", "", 5800 - 135};
> RooArgusBG argus{"argus", "", obs, argus\_m0, argus\_par};
> RooGaussian resolution{"resolution", "", obs, RooFit::RooConst(0), sigma};
> RooFFTConvPdf conv{"conv", "", obs, argus, resolution};
> 
> RooAddPdf model{"model", "", {sigpdf, conv}, {RooConst(0.3)}};
> 
> std::unique\_ptr\<RooDataSet\> data{model.generate(obs, 1000)};
> 
> std::unique\_ptr\<RooFitResult\> fit\_result{model.fitTo(\*data, Save(true))};
> 
> // This works!
> // std::unique\_ptr\<RooFitResult\> fit\_result{model-\>fitTo(\*data, Save(true), EvalBackend("legacy"))};
> 
> fit\_result-\>Print("v");
> \`\`\`
> The output is:
> \`\`\`txt
> RooFitResult: minimized FCN value: 1.1609e-12, estimated distance to minimum: 5.82024e-16
> covariance matrix quality: Not calculated at all
> Status : MINIMIZE=-1 HESSE=302
> 
> Constant Parameter Value
> -------------------- ------------
> argus\_m0 5.6650e+03
> argus\_par -5.5000e+01
> 
> Floating Parameter InitialValue FinalValue +/- Error GblCorr.
> -------------------- ------------ -------------------------- --------
> mean 5.8000e+03 5.8000e+03 +/- 0.00e+00 \<none\>
> sigma 8.1000e+00 8.0336e+00 +/- 0.00e+00 \<none\>
> \`\`\`
> With the \`"legacy"\` evaluation backend, the fit succeeds and both the \`MINIMIZE\` and \`HESSE\` status codes are zero, indicating that there is no error.
