Java Reference
In-Depth Information
10.
return;
11.
}
12.
int lSelectedFunction
mFunctionPane.getSelectedIndex();
13.
String lFunctionNameShort
"";
14.
BuildSettings lBuildSettings
null;
In line 3, we set the cursor to indicate to the user that the model
build execution time can be long. We reuse the
createDatasetAndIm-
portMetaData
method to ensure the
mBuildData
is aligned with the
last dataset URI entered by the user and that the dataset metadata
has been retrieved. The member
mKeyForAssociation
contains the
name of the attribute selected by the user as a key, which will be used
when creating the specific mining function
BuildSettings
. Line 8
shows that we use the check box value to ask the DME to overwrite
objects already existing in the MOR. We get back the selected mining
function through the associated selected pane (or tab). In line 13, we
prepare a string to contain a short prefix depending on the user-
selected mining function. This prefix will be used to generate names
of objects, such as
BuildSettings
or
PhysicalDataSet,
to be saved.
15.
switch (lSelectedFunction) {
16.
//Classification
17.
case CLASSIFINDEX :
lFunctionNameShort
"Cla";
18.
lBuildSettings
createClassificationSettings(
19.
20.
(String) mTargetNameClassif.getSelectedItem());
21.
break;
We extract the user-selected target attribute name from the associ-
ated member variable pointing to a combo box. This is the only infor-
mation needed for classification. The section for regression follows.
22.
case REGINDEX :
23.
lFunctionNameShort
"Reg";
24.
lBuildSettings
createRegressionSettings(
25.
(String) mTargetNameRegression.getSelectedItem());
26.
break;
We extract the user selected target attribute name from the associ-
ated member variable pointing to a combo box. This is the only
needed information for regression.
Search WWH ::
Custom Search