Options
"Read CSV defaults column names to gensym*"
awchisholm
RapidMiner Certified Expert, Member Posts: 458 
Unicorn
Hello all,
If I read a CSV file where the first data row does not contain the attribute names, I notice that all regular attributes are given constructed names like gensym, gensym1 and so on.
If I try to edit the meta data information to specify the attribute names, I can't find a way to get the names I want to be used. If I change the type of the attribute to be a special type, it imports with the correct name. In this case, I notice that sometimes, two duplicate columns are created, one regular with the name gensym and another special type with the correct name.
Obviously I can work round it but I think the behaviour has changed in the latest release (5.2.1) because I had a process that used to work that failed after the upgrade.
regards
Andrew
If I read a CSV file where the first data row does not contain the attribute names, I notice that all regular attributes are given constructed names like gensym, gensym1 and so on.
If I try to edit the meta data information to specify the attribute names, I can't find a way to get the names I want to be used. If I change the type of the attribute to be a special type, it imports with the correct name. In this case, I notice that sometimes, two duplicate columns are created, one regular with the name gensym and another special type with the correct name.
Obviously I can work round it but I think the behaviour has changed in the latest release (5.2.1) because I had a process that used to work that failed after the upgrade.
regards
Andrew
Tagged:
0
Answers
that is most probably related to the fix of this thread and slipped through our testing procedure. We will fix it for the next release. The problem should not occur in 5.2.0.
Best,
Marius