Please have a look at the screen shot. On the connection tab,all tables from Administrator schema are listed, but when I try to drag to the MiningFlow or use the node Table Source, it does not work nor does list two tables. But I can see the other three tables in the same schema.
Secondly, I am using the predictor---Tree approach. Which splitting criteria does this algorithm uses gini index or entropy or something else?
NOTICE: developerWorks Community will be offline May 29-30, 2015 while we upgrade to the latest version of IBM Connections. For more information, read our upgrade FAQ.
This topic has been locked.
2 replies Latest Post - 2011-04-20T14:50:31Z by Nafeespari
Pinned topic Tables Source does not list all tables in MiningFlow
Answered question This question has been answered.
Unanswered question This question has not been answered yet.
Updated on 2011-04-20T14:50:31Z at 2011-04-20T14:50:31Z by Nafeespari
SystemAdmin 110000D4XK203 PostsACCEPTED ANSWER
Re: Tables Source does not list all tables in MiningFlow2011-04-20T14:45:04Z in response to NafeespariHi,
The mining flow reads database metadata from a database model file (.dbm), it does not read it directly from the database. If the model file is not up-to-date, you will see differences between the tree in the Data Source Explorer and the tree in the Table selection dialog. To update the database model, select Mining Flow -> Refresh Database in the main menu. This is needed when you create/delete tables, views etc.
The Tree algorithm uses the gini index.