"_LOCAL Repository not found while reading CSV file"

SC

Hello,

I am new to Rapid Miner/Analytics. As a first step to the data mining exercise, I am trying to read an xlsx file(setting in input) hosted on the Rapid Analytics Server under my account /home/xxxxx/examples/10000005.xlsx. I exported the process as a service and I am able to access it using the public_access URL for RA. However, when I log out from RA server, I get this error:
<type>de.rapidanalytics.entity.RepositoryException</type><message>Error executing process /home/XXXX/getPackageList for service getPackageList: Cannot retrieve repository data from entry '//_LOCAL/home/XXXX/10000005.xlsx'. Reason: Requested repository _LOCAL does not exist..</message>

Please help, it took me 2 mins to build the process/service but 1 day to figure out how can my devs use it.

Thanks so much!

Marco_Boeck

Hi,

you need to differentiate between a repository (the storage place for RapidMiner data, e.g.example sets, models, etc) and the normal filesystem. Store/Retrieve operators work on the RapidMiner repository, while the Import operators (like "Read Excel") work on the filesystem. The usual workflow is to import data from the filesystem into the RapidMiner repository (which converts them to a so called "example set", the format RapidMiner is using) and then work with the data from the repository.

Regards,
Marco

elnecesario

Hi,

is it possible for operators like "Read Excel" to use environmental variables or macros for a relative path? If not, is there another possibility, e.g. a path relative to the process-location?

Regards,
elnecesario

Marco_Boeck

Hi,

yes you can use macros in the path. A relative filesystem path is not available however, you do need a macro specifying a fully qualified path to dynamically change the file read.

Regards,
Marco