Exclamation mark in "Reterive"

pemguinkplpemguinkpl Member Posts: 14 Contributor II
Hai,

i still new in rapid miner, may i ask
why my data have Exclamation mark???
i not understand what the means for
"parameter repository entry accesses a repository by name(//H1N1/test/2). This may not be portable when sharing processes."

anyone give some comment and help...thank you!!!

Answers

  • MariusHelfMariusHelf RapidMiner Certified Expert, Member Posts: 1,869 Unicorn
    Hi, it just means that you are using an absolute path. If your process is stored in the same repository as your data, you should use relative paths. The easiest way to do this is to reselect the data entry by clicking the button in the operator parameters, and check "Resolove relative to..." in the dialog.
    If your processes and data are stored in different repositories (which we do not recommend) you can savely ignore the warning.

    Regards,
    Marius
  • pemguinkplpemguinkpl Member Posts: 14 Contributor II
    thank you Marius  ;)
  • dranammaridranammari Member Posts: 13 Contributor II
    Hi Marius,

    I am having the same message in two "store" operators. One is used to store the wordlist from a "Process Documents" operator, and the other is used to store a model from a "X-Validation" operator. My data is being read from a mySQL database using the "Read Database" operator. Now I am developing this process on my laptop using the RapidMiner GUI. What I want is to integrate this process into a Java application (using the RapidMiner API) and then move the Java application on a server and be able to run the application (with the integrated process).

    My problem is that when I move the Java application, the absolute paths in the "store" operators will not exist on the server and most likely I will get an error. How can I address this problem?

    Many thanks,
    Ahmad 
    Marius wrote:

    Hi, it just means that you are using an absolute path. If your process is stored in the same repository as your data, you should use relative paths. The easiest way to do this is to reselect the data entry by clicking the button in the operator parameters, and check "Resolove relative to..." in the dialog.
    If your processes and data are stored in different repositories (which we do not recommend) you can savely ignore the warning.

    Regards,
    Marius
  • Marco_BoeckMarco_Boeck Administrator, Moderator, Employee, Member, University Professor Posts: 1,993 RM Engineering
    Hi,

    very simple solution: Use RapidAnalytics ;)

    Regards,
    Marco
  • MariusHelfMariusHelf RapidMiner Certified Expert, Member Posts: 1,869 Unicorn
    Basically, Marco is right - RapidAnalytics is perfectly fit for your problem. It supplies you with a so called remote repository, which for you looks the same as a standard local repository inside rapidminer, but your data and processes are actually stored on a RapidAnalytics server. If you have your data in that repository, you can easily (with 2 clicks) run the process on the server or on your local machine, just as you prefer.
    You can also export your processes as a webservice and query them via http requests.

    If you don't want to setup a RapidAnalytics instance, just make sure that the repositories on your local RapidMiner instance and the server instance have the same structure. If you use relative paths, you should be able to run your processes both on the laptop and on the server.

    Btw, you can use relative paths in the store operators the same way as with the retrieve operators - just remove the leading absolute path prefix in the input field of the store operator.

    Best,
    Marius
Sign In or Register to comment.