I have an ArrayList full of these:
class TransitionState {
Position positionA;
Position positionB;
int counter;
public boolean equals (Object o){
if (o instanceof TransitionState){
TransitionState transitionState= (TransitionState)o;
if ((this.positionA.equals(transitionState.positionA))
&&(this.positionB.equals(transitionState.positionB)))
{
return true;
}
}
return false;
}
@Override
public String toString() {
String output = "Position A " + positionA.i+ " "+ positionA.j + " "+ positionA.orientation + " "+
"Position B " + positionB.i + " "+ positionB.j + " "+ positionB.orientation;
return output;
}
}
class Position {
int i;
int j;
char orientation;
Position() {
}
void setIJ(int i, int j){
this.i=i;
this.j=j;
}
void setOrientation(char c){
orientation = c;
}
public boolean equals(Object o){
if(o instanceof Position){
Position p = (Position)o;
if((p.i==this.i)&&(p.j==this.j)&&(p.orientation==this.orientation))
{
return true;
}
else return false;
}
return false;
}
} //end class Position
I query it with this:
if(!transitionStatesArray.contains(newTransitionState)){ //if the transition state is new add and enqueue new robot positions
transitionStatesArray.add(newTransitionState); //marks as visited
I'm finding duplicate elements inside my transitionStatesArray
, why is this?
I'm using these i,j and orientation values to fill unique values in a matrix, yet here I have a duplicate:
S . N
* * *
. D D
E . O
* * *
. D D
N . S
* * *
. D D
S . N
* * *
. D D
The List.contains(...)
method is defined to use equals(Object)
to decide if the argument object is "contained" by the list. So you need to override equals
... assuming that the default implementation is not what you need.
However, you need to be aware that List.contains(...)
potentially tests the argument against every element in the list. For a long list, that's expensive. Depending on the details of your application, it may be better to use a different collection type (e.g. a HashSet
, TreeSet
or LinkedHashSet
) instead of a List
. If you use one of those, your class will need to override hashCode
or implement Comparable
, or you will need to create a separate Comparator
... depending on what you choose.
(A bit more advice on the alternatives ... since the OP is interested)
The performance of contains
on a List
type like ArrayList
or LinkedList
is O(N)
. The worst-case cost of a contains
call is directly proportional to the list length.
For a TreeSet
the worst-case performance of contains
is proportional to log2(N)
.
For a HashSet
or LinkedHashSet
, the average performance of contains
is a constant, independent of the size of the collection, but the worst-case performance is O(N)
. (The worst case performance occurs if you 1) implement a poor hashcode()
function that hashes everything to a small number of values, or 2) tweak the "load factor" parameter so that the hash table doesn't automatically resize as it grows.)
The downside of using Set
classes are:
get(pos)
method, andSet
classes don't even preserve the insertion order.These issues need to be considered when deciding what collection class to use.