Develop Reference

Weka J48 classification not following tree

My original tree was much bigger, but since I'm stuck with this issue for quite some time I decided to try to simplify my tree. I Ended up with something like this:
As you can see, I only have a single attribute called "LarguraBandaRede" with 3 possible nominal values "Congestionado", "Livre" and "Merda".
After that I exported the j48.model from weka to use on my java code.
With this piece of code I import the model to use as a classifier:
ObjectInputStream objectInputStream = new ObjectInputStream(in);
classifier = (J48) objectInputStream.readObject();
After that I started to create a arraylist of my attributes and a Instances File
for (int i = 0; i <features.length; i++) {
String feature = features[i];
Attribute attribute;
if (feature.equals("TamanhoDados(Kb)")) {
attribute = new Attribute(feature);
} else {
String[] strings = null;
if(i==0) strings = populateAttributes(7);
if(i==1) strings = populateAttributes(10);
ArrayList<String> attValues = new ArrayList<String>(Arrays.asList(strings));
attribute = new Attribute(feature,attValues);
}
atts.add(attribute);
}
where populateAttributes gives the possible values for each attribute, in this case "Livre, Merda, Congestionado;" for LarguraBandaRede and "Sim,Nao" for Resultado, my class attribute.
Instances instances = new Instances("header",atts,atts.size());
instances.setClassIndex(instances.numAttributes()-1);
After creating my instances is time to create my Instance File, that is, the instances that I'm trying to classify
Instance instanceLivre = new DenseInstance(features.length);
Instance instanceMediano = new DenseInstance(features.length);
Instance instanceCongestionado = new DenseInstance(features.length);
instanceLivre.setDataset(instances);
instanceMediano.setDataset(instances);
instanceCongestionado.setDataset(instances);
then I set each of this instances with the 3 possible values for "LarguraBandaRede". 'instanceLivre' with "Livre", 'instanceMediano' with "Merda" and 'instanceCongestionado' with "Congestionado".
After that I only classify this 3 instances using the classifyInstance method
System.out.println(instance.toString());
double resp = classifier.classifyInstance(instance);
System.out.println("valor: "+resp);
and this is my result:
As you can see, the instance that has Merda as "LarguraBandaRede" was classify to be the same class as Congestionado, the class 'Nao'. But that doesn't make any sense, since the tree above clearly show that when "LarguraBandaRede" is "Merda" or "Livre" the class should be the same.
So that's my question. How this happened and how to fix it?
Thanks in advance.
EDIT
I didn't know that this:
made any difference in the way the model works. But we have to follow this order when feeding a nominal attribute with possible values.

Have you checked if the weka nominal attribute index is equal in order to your populateAttributes method?

Invoke Methods Dynamically on Java

At work, we have to generate a report for our client that changes its parameters several times during the week.
This report is generated from a single table on our database.
For example, imagine a table that has 100 columns and I have to generate a report with only 5 columns today, but tomorrow I have to generate with 95 of them.
With this in mind, I created a TO class with all the columns of the specified table and my query returns all columns (SELECT * FROM TABLE).
What I'm trying to create is a dynamic form to generate the report.
I first thought on create a simple frame with a list of the columns listed as check boxes and the user would select the columns that he wants (of course with a button to Select All and another to Deselect All).
As all of the columns have the same name as the attributes of the TO class, I developed the following code (I have Google this):
Class c = Test.class;
for(int i = 0; i < listOfAttributes.length; i++)
{
auxText += String.valueOf( c.getMethod( "get" + listOfAttributes[i]).invoke( this, null ) );
}
Is this the better way to do what I need to?
Thanks in advance.
Obs.: the getters of the TO class have the pattern "getAttribute_Name".
Note: This question is different from the one where the user is asking HOW to invoke some method given a certain name. I know how to do that. What I'm asking is if this is the better way to solve the problem I described.

My Java is a little more limited, but I believe that's about as good as you're going to get using reflection.
Class<?> c = Test.class;
for (String attribute : listOfAttributes) {
auxText += String.valueOf(c.getMethod("get" + attribute).invoke(this, null));
}
But since this sounds like it's from potentially untrusted data, I would advise using a HashMap in this case, with each method explicitly referenced. First of all, it explicitly states what methods can be dynamically called. Second, it's more type safe, and compile-time errors are way better than runtime errors. Third, it is likely faster, since it avoids reflection altogether. Something to the effect of this:
private static final HashMap<String, Supplier<Object>> methods = new HashMap<>();
// Initialize all the methods.
static {
methods.set("Foo", Test::getFoo);
methods.set("Bar", Test::getBar);
// etc.
}
private String invokeGetter(String name) {
if (methods.containsKey(name)) {
return String.valueOf(methods.get(name).get());
} else {
throw new NoSuchMethodException();
}
}
It might sound like a major DRY violation to do so, but the repetition at least makes sure you don't wind up with unrelated getters accidentally called.

Class c = Test.class;
for(int i = 0; i < listOfAttributes.length; i++)
{
auxText += String.valueOf( c.getMethod( "get" + listOfAttributes[i]).invoke( this, null ) );
}
You can do this somewhat more elegantly via Java Beans, the Introspector, and PropertyDescriptor, but it's a little more long-winded:
Map<String, Method> methods = new HashMap<>();
Class c = this.getClass(); // surely?
for (PropertyDescriptor pd : Introspector.getBeanInfo(c).getPropertyDescriptors())
{
map.put(pd.getName(), pd.getReadMethod();
}
//
for (int i = 0; i < listOfAttributes.length; i++)
{
Method m = methods.get(listOfAttributes[i]);
if (m == null)
continue;
auxText += String.valueOf(m.invoke(this, null));
}

How to say String A > String B in Java

I'm working on a small program to compare service levels, the user will input the service level 2 times (current and requested) and then the input will be scanned and compared and show a message.
For example:
current = 9*5 NBD (a)
requested = 24*7 SBD (b)
I want to know how in Java I can tell the compiler that (b) is greater than (a)
Because I want to use if statement like this
if (b > a) then show message.
I tried to use string.equals, but didn't help me too much.
I was not successful to convert string to number to do such comparison.

Try following statement
if(a.compareTo(b) > 0);
First thing: you can't override String.compareTo(), because it's final. you can create class with String field and write compareTo() for this class. This is not best idea.
But you can compare two strings by putting them into array and creating implementation of Comparator interface in sort() method.
String current = "9*5 NBD";
String requested = "24*7 SBD";
String[] test = {current, requested};
Arrays.sort(test, new Comparator<String>() {
#Override
public int compare(String o1, String o2) {
//Your impl goes here
return 0;
}
});

Where do these strings come from? Surely they must be from some kind of table that contains the service level details together with the cost of subscribing to that service level. What you want to check is whether the required service level costs more than the service level the client already has. Suppose the service level details come from a Map<ServiceLevel, BigDecimal> that gives the cost for a certain service level. Then all you need to do is:
BigDecimal costOfCurrentSL = serviceLevelCosts.get(currentSL);
BigDecimal costOfRequiredSL = serviceLevelCosts.get(requiredSL);
if (costOfRequiredSL.compareTo(costOfCurrentSL) > 0) {
// ... tell client he needs to purchase a top-up
}

Thank you all for the willing to help :)
After a lot of thinking I found another way which helped me a lot
I created 2 new integers and called them Values of what I need.
and used if statement, that if the entered is 9*5 NBD so the value will be zero, and if it is SBD, the value will be 1 and so on, then created new if statement to compare the values and show me a message if the A is greater than B, and it really worked.
Here is a part of my code
String WA_SLA = "", REQ_SLA = "";
int Va_WA_SLA = 0, Va_REQ_SLA = 0;
if(WA_SLA.equalsIgnoreCase("9*5 SBD"))
{
Va_WA_SLA = 1;
}
if(REQ_SLA.equalsIgnoreCase("9*5 NBD"))
{
Va_REQ_SLA = 0;
}
if(Va_WA_SLA > Va_REQ_SLA)
{
JOptionPane.showMessageDialog(frame,"Warranty SLA is Higher than Requested SLA " ,null, JOptionPane.INFORMATION_MESSAGE);
}
Thaaaaaaaaaaaaank you a lot

I fail junit tests because of return type (I think)

So I have a class full of junit tests and a class full of methods that perform binary operations. The tests are checking to see if I have the right values at certain points.
I am failing a lot of tests because of what I believe to be is the return type. For example I get the message
junit.framework.ComparisonFailure: null expected:<[000]> but was <[BinaryNumber#4896b555]>
If I'm understanding this it's saying that it was looking for an array containing 000 but it got a BinaryNumber (which is the required return type). To help clarify here is one of the methods.
public BinaryADT and(BinaryADT y) {
int[] homeArr = pad(((BinaryNumber) y).getNumber());
int[] awayArr = ((BinaryNumber) y).pad(getNumber());
int[] solution = new int[awayArr.length];
int i = 0;
String empty = "";
while(i < solution.length){
solution[i] = homeArr[i] & awayArr[i];
i++;
}
for(int indexF = 0; indexF < solution.length; indexF++){
empty = empty + solution[indexF];
}
System.out.println(empty);
return new BinaryNumber(empty);
}
Am I understanding this right? If not could someone please explain? I'd also like to point out that this is for my homework but I'm not asking for answers/someone to do it for me. Just a point in the right direction at most.
I will gladly clarify more if it is needed (I didn't want to bog everything down).
Also this is my first post on here. I tried to keep to the formatting suggestions but I apologize if anything is sub-par.
As suggested here is the test method
public void testAnd1()
{
BinaryADT x = new BinaryNumber("111");
BinaryADT y = new BinaryNumber("000");
BinaryADT z = x.and(y);
assertNotSame(x,z);
assertNotSame(y,z);
assertEquals("000",z.toString());
}

Whenever you see the output of "toString()" like ClassName#SomeNumber, then you can be sure that toString() method is not implemented for that class (unless toString() method implementation itself is not like this).
In your case, expected value is [000], but you are getting [BinaryNumber#4896b555].
Try to implement toString() method in BinaryNumber class and return the value from this method as per assertEquals() expects. This should solve the problem.

Can you show me your test code?
1.Your expected type is different from the actual type.
2.BinaryADT class didn't overide toString method.

Reading Object in List; cannot be cast

I think I have a simple mistake in my code but I can't find it.
I have a list of Objects (type of an entity) and I want to read the content of the objects in the list.
In my opinion something like:
object.get(1).getTitle();
List<HtMeldungen> meldungen = q.getResultList();
List<MeldungsBean> meldungsliste = new ArrayList();
MeldungsBean mb = null;
HtMeldungen tempMeldungen = null;
int i = 0;
int k = meldungen.size() - 1;
for (i = 0; i < k; i++) {
mb = new MeldungsBean();
tempMeldungen = (HtMeldungen) meldungen.get(i);
mb.setTitel(tempMeldungen.getTitle());
mb.setAutor(tempMeldungen.getAutor());
mb.setMeldungstext(tempMeldungen.getText());
meldungsliste.add(mb);
}
My list named meldungen is filled with objects of type HtMeldungen.
I get the error:
DBEntities.classic.HtMeldungen cannot be cast to DBEntities.classic.HtMeldungen
Can anyone help me?

Are you sure q.getResultList() gets a list with instances of HtMeldungen?
If not, then the line
List<HtMeldungen> meldungen = q.getResultList();
is - depending of your compiler switches - syntactically correct, but the list can contain instances of a different class, and later in the line
tempMeldungen = (HtMeldungen) meldungen.get(i);
you get your exception, because that what the compiler thinks it must be instance of HtMeldungen in fact isn't.
Try the code
if (meldungen.get(i) instanceof HtMeldungen) {
tempMeldungen = (HtMeldungen) meldungen.get(i);
} else {
throw new RuntimeException("Got instance of class " + meldungen.get(i).getClass());
}
then you get an understandable error if your assumption should have been wrong.

I'll get the error: DBEntities.classic.HtMeldungen cannot be cast to DBEntities.classic.HtMeldungen
Since the error message indicates that an object of HtMeldungen cannot be cast to HtMeldungen (which seems contradictory), I would think that you might have this class loading twice in your build. Please check to see if your build path is putting the same jar in the build twice. That is what usually causes this error.