Question:
Learning Java 8, I discovered streams. I decided to try to parallelize long enough nested loops, but I ran into a problem – I don’t know how yet.
This is what the sequential code looks like:
long total = 0;
BitVector vector = getVector(); // из библиотеки import cern.colt.bitvector.*
for (int index = 0; index < vector.size(); index++){
if (vector.get(index)){
for (TrafficData trafficData : trafficDataList){
int visitorTrafficDataIndex = getVisitorTrafficDataIndex(index, trafficData);
if (visitorTrafficDataIndex >= 0){
totalTrafficImpressions += (long)(trafficData.getVolume()[visitorTrafficDataIndex]);
}
}
}
}
I thought of doing something like this:
long total = IntStream.range(0, vector.size()).parallel()
.filter(index -> vector.get(index))
.mapToLong(index ->
trafficDataList.stream()
.map(trafficData -> getVisitorTrafficDataIndex(index, trafficData))
.filter(visitorTrafficDataIndex -> visitorTrafficDataIndex >= 0)
.mapToLong(visitorTrafficDataIndex -> (long)(trafficData.getVolume()[visitorTrafficDataIndex]) // здесь уже нет объекта использовать trafficData
.sum())
).sum();
The problem is it doesn't compile ..
Answer:
If I understand your problem correctly, then you need to do something like this:
long total = IntStream.range( 0, vector.size() ).parallel().filter( index -> vector.get( index ) )
.mapToLong( index -> trafficDataList.stream().parallel().map( trafficData -> {
int visitorTrafficDataIndex = getVisitorTrafficDataIndex( index, trafficData );
if (visitorTrafficDataIndex >= 0) {
return trafficData.getVolume()[visitorTrafficDataIndex];
}
return 0;
} ).sum() ).sum();
You won't be able to do it the way you want (everything on streams), because after you map(trafficData -> getVisitorTrafficDataIndex(index, trafficData))
there is no trafficData
object in your trafficData
, it (stream) is just a collection of int.