superslow sys_calc in OS.x

No I’m not aware of such problem, it works perfectly fine for me.

one guess is that you comment the line:
> #matchscale:
> #32 50 80

This might be creating trouble and might put the code in a infinite loop.
Could you retry without commenting those lines?

Cheers,

Olivier

On 05 Feb 2015, at 00:02, Valerio Dao <email address hidden> wrote:

> New question #261689 on MadGraph5_aMC@NLO:
> https://answers.launchpad.net/mg5amcnlo/+question/261689
>
> Hello,
>
> I am experiencing some trouble in running sys_calc on OSx (Mevericks).
>
> If I try to run it during the normal MG generation I get the message:
> SysCalc Failed. Please read the associate log to see the reason. Did you install the associate PDF set?
> ... more specifically:
> Init PDF set NNPDF23_lo.LHgrid
> libc++abi.dylib: terminating with uncaught exception of type LHAPDF::UserError: Can't find a valid PDF NNPDF23_lo/0
>
> Now given that I am not able to correctly use the LHAPDF installation within mad graph
> (see other tread concerning obscure compiler problems not solved since october), I tried to run as a post processing.
>
> I use the following config.txt:
> scalefact:
> 0.5 1 2
> scalecorrelation:
> -2
> alpsfact:
> 0.5 1 2
> # matching scales
> #matchscale:
> #32 50 80
> # PDF sets and number of members (optional)
> PDF:
> CT10.LHgrid 10
>
> and then do:
> ./sys_calc ../QCD_emu/Events/run_02/unweighted_events.lhe.gz config.txt test.lhe
>
> the input lhe file has 1k events.
>
> What happens is that soon after the start:
> Init PDF set CT10.LHgrid
> LHAPDF 6.1.3 loading /opt/local/share/LHAPDF/CT10/CT10_0000.dat
> CT10 PDF set, member #0, version 4; LHAPDF ID = 10800
> Using 10 members for this set
>
> the CPU consumption goes to 100% and my laptop becomes unusable.
> I let it run for 10 minutes but nothing has moved since then (I would expect the code to do better than 2ev/sec).
>
> Is this a known issue?
>
> thanks a lot,
>
> Valerio
>
>
>
> --
> You received this question notification because you are an answer
> contact for MadGraph5_aMC@NLO.

Thanks for the prompt answer.

Leaving in those lines didn't solve the problem.
What actually did solve the problem was to pass an untarred lhe file so:

unweighted_events.lhe.gz

instead of

unweighted_events.lhe

Cheers,

Valerio