To prevent spam users, you can only post on this forum after registration, which is by invitation. If you want to post on the forum, please send me a mail (h DOT m DOT w DOT verbeek AT tue DOT nl) and I'll send you an invitation in return for an account.

CSV to XES seems to have an error in the ordering of events within traces

I have imported the attached .txt file into ProM 6.4.1 and created a Log (via Convert Key/Value Set to Log), and then used the resultant Log to create a visual graph using the Mine with Inductive Visual Miner.  What I got seems to have a discrepancy in the process flow since it shows one path bypassing the 'Create' step (Graph_1.png).  An examination of the .xes file of the Log shows what may be the reason for the issue. It appears that the Create event for Emily Knight is being loaded at the end of trace 1 (Joseph Whitehouse) instead of at the beginning of trace 2 (Emily Knight).  If I edit the .xes to fix this error the resulting Petri net graph is as I would have expected (Graph_2.png).  Am I correct in this analysis? And if so, has this error been reported?







Answers

  • Dear XCalRocketMan,

    Thanks for reporting this issue. I believe this has been reported and fixed, as ProM 6.4.1. was released 2 years ago. Please try ProM lite or ProM 6.6 from promtools.org and verify if the issue has been resolved. If not, please let us know!!!
    Joos Buijs

    Senior Data Scientist and process mining expert at APG (Dutch pension fund executor).
    Previously Assistant Professor in Process Mining at Eindhoven University of Technology
  • I will try ProM 6.6, thanks.  I'll let you know if it was fixed in that version.

    NOTE: the reason I am using 6.4.1 is that I am trying to build some custom code to invoke ProM resources via the ProM Command Line Interface/API as documented marginally in a blog: https://dirksmetric.wordpress.com/2015/03/11/tutorial-automating-process-mining-with-proms-command-line-interface/

    I believe when a colleague tried the CLI on the latest version of ProM (6.6) it didn't work.

    Is there documentation available that describes the use of the CLI/API for 6.6?  And if so, where?  I have not been able to locate any comprehensive documentation on the CLI for either 6.4.1 or 6.6.

    Thanks in advance,

    Michael

  • Dear Michael,

    could you please try the CLI on ProM 6.6 and tell where it fails? Then we can look into this. We did not remove the CLI from ProM. Some plugins might have changed. The Blog post for now is the only more extensive documentation. But it's basically also all technical information we can provide.

    Each script is full Java-code that gets interpreted at run-time. The only addition are the names of the ProM plugins to invoke, which can be retrieved via the "-l" command-line switch.

    Related to this: ProM 6.6 (and up) contains a Plugin to reorder events in a log based on time-stamps.

    Kind regards,
    Dirk

  • Dirk,

    Thanks for the reply and the information.  It turns out that the error in the graphing was 'my bad' ... I had an incorrect value in the original csv file that caused ProM to create an unexpected graph.  Turns out that ProM will, in fact, provide a proper graph if I send it proper data :)  

    Anyway, thanks again.

    With regard to the CLI on 6.6, I'll need to run some additional test cases later this week (or next) and get back to you. 

    Michael
Sign In or Register to comment.