Calculating average of an array list?

Why use a clumsy for loop with an index when you have the enhanced for loop?

private double calculateAverage(List <Integer> marks) {
  Integer sum = 0;
  if(!marks.isEmpty()) {
    for (Integer mark : marks) {
        sum += mark;
    }
    return sum.doubleValue() / marks.size();
  }
  return sum;
}

Update: As several others have already pointed out, this becomes much simpler using Streams with Java 8 and up:

private double calculateAverage(List <Integer> marks) {
    return marks.stream()
                .mapToDouble(d -> d)
                .average()
                .orElse(0.0)
}

Use a double for the sum, otherwise you are doing an integer division and you won't get any decimals:

private double calculateAverage(List <Integer> marks) {
    if (marks == null || marks.isEmpty()) {
        return 0;
    }

    double sum = 0;
    for (Integer mark : marks) {
        sum += mark;
    }

    return sum / marks.size();
}

or using the Java 8 stream API:

    return marks.stream().mapToInt(i -> i).average().orElse(0);

From Java8 onward you can get the average of the values from a List as follows:

    List<Integer> intList = Arrays.asList(1,2,2,3,1,5);

    Double average = intList.stream().mapToInt(val -> val).average().orElse(0.0);

This has the advantage of having no moving parts. It can be easily adapted to work with a List of other types of object by changing the map method call.

For example with Doubles:

    List<Double> dblList = Arrays.asList(1.1,2.1,2.2,3.1,1.5,5.3);
    Double average = dblList.stream().mapToDouble(val -> val).average().orElse(0.0);

NB. mapToDouble is required because it returns a DoubleStream which has an average method, while using map does not.

or BigDecimals:

@Test
public void bigDecimalListAveragedCorrectly() {
    List<BigDecimal> bdList = Arrays.asList(valueOf(1.1),valueOf(2.1),valueOf(2.2),valueOf(3.1),valueOf(1.5),valueOf(5.3));
    Double average = bdList.stream().mapToDouble(BigDecimal::doubleValue).average().orElse(0.0);
    assertEquals(2.55, average, 0.000001);
}

using orElse(0.0) removes problems with the Optional object returned from the average being 'not present'.

With Java 8 it is a bit easier:

OptionalDouble average = marks
            .stream()
            .mapToDouble(a -> a)
            .average();

Thus your average value is average.getAsDouble()

return average.isPresent() ? average.getAsDouble() : 0;