[Rivet] Problem with multiprocessing in rivet 2.4.1

[Holger Schulz]

Hi David,

yeah this has to do with the .plot file of the analysis.

It affects only the d01-x01-y01 plot.

The XCustomMajorTicks seems to be broken.

To exclude that one from being plotted, just add this to your
rivet-mkhtml command line:

    -M "\/ATLAS_2014_I1319490\/d01-x01-y01"

To plot it it's probably easiest to delete the XCustomMajorTicks
statement from ATLAS_2014_I1319490.plot in data/plotinfo
and do make install in there.

Best,
Holger


On 31/03/16 10:19, Yallup, David wrote:
>
> Hi,
>
>
> I updated to rivet 2.4.1 and yoda 1.5.9 and was generating events for 
> the analysis ATLAS_2014_I1319490,
>
>
> The yoda file appears to have created correctly (attached a copy with 
> basically no events), but I get the following error when running 
> rivet-mkhtml on it
>
>
> david at yallup:~/Desktop/WeakTest$ rivet-mkhtml LHC.yoda
> Making 34 plots
> Traceback (most recent call last):
>   File "/home/david/Install/Herwig-7.0.0/bin/make-plots", line 2607, 
> in <module>
>     for i, _ in enumerate(pool.imap(process_datfile, datfiles)):
>   File "/usr/lib/python2.7/multiprocessing/pool.py", line 659, in next
>     raise value
> ValueError: invalid literal for float(): 0$
>
>
> Couple of notes:
>
> - It is just for this analysis routine (I suspect as this is one with 
> selectable modes for muon and e channels)
>
> - It runs fine if I swap the make-plots from 2.4.1 out for the 2.4.0 
> (think this multiprocessing was new)
>
>
> Seemed like a bug to me, but I might have messed up along the way 
> somewhere!
>
> Thanks for your help,
>
> David
>
>
>
> _______________________________________________
> Rivet mailing list
> Rivet at projects.hepforge.org
> https://www.hepforge.org/lists/listinfo/rivet

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.hepforge.org/lists-archive/rivet/attachments/20160331/e9d7b91e/attachment.html>