How can I filter rows having particular column value which is above average using java for input comma separated file

Question

Suppose I have some file having some data in comma separated format as below

TIMESTAMP,COUNTRYCODE,RESPONSETIME
   1544190995,US,500
   1723922044,GB,370
   1711557214,US,750

How can I filter rows by RESPONSETIME above average using java ?ie Here the average of RESPONSETIME is 526.So I need to display all rows having RESPONSETIME greater than 526.Here datalines are not guaranteed to be in any particular order.Can we do both(finding average and filter rows having RESPONSETIME above average) in a single method?

Currently I am finding average as below .How can I apply filter and return as collection inside the same method?As per my understanding its not possible to read a file twice inside same method.

        public static Collection<?> filterByResponseTimeAboveAverage(Reader source) {
            BufferedReader br = new BufferedReader(source);
            String line = null;
            Collection<String> additionalList = new ArrayList<String>();
            int iteration = 0;
            String[] myArray = null;
            long count=0;
            long responseTime=0;
            long sum=0;
            int numOfResponseTime=0;
            long average=0;
            List<String> myList = new ArrayList<String>();
            try
            {
                while ((line = br.readLine()) != null) {
                    System.out.println("Inside while");
                    if (iteration == 0) {
                        iteration++;
                        continue;
                    }
                    myArray = line.split(",");
                    for (String eachval:myArray)
                    {

                        boolean isNumeric = eachval.chars().allMatch(x -> Character.isDigit(x));
//since input dataline is not guaranted to be in any particular order I am finding RESPONSETIME like this
                        if (isNumeric)
                        {
                        count=eachval.chars().count();

                        if (count<10)
                        {
                            responseTime=Integer.parseInt(eachval);
                            sum=sum+responseTime;
                            numOfResponseTime++;
                        }
                    }
                        myList.add(eachval);
                    }

                }
                    average=sum/numOfResponseTime;
                    System.out.println("Average -- "+average);
                      ---------------
                      ---------------
    }

Answer 1

How can I apply filter and return as collection inside the same method?

Use Java 8 streams and lambdas

Answer 2

As per my understanding its not possible to read a file twice inside same method.

You can but you should not do this as it is not efficient.

---

You have mainly two ways of proceeding.

Optimized way :

• reading all values from the file and compute the average of RESPONSETIME.
• filtering values above the average

You could introduce a private methods invoked by filterByResponseTimeAboveAverage() to retrieve both all values from the source and compute average of them.

Functional way (a little more expensive in overhead) :

• reading all values from the file
• use IntStream.average() to compute the average of RESPONSETIME.
• filtering values above the average

For the second and last step, it could be like :

double average = list
                .stream()             
                .mapToInt(MyObject::getAverage)
                .average()
                .getAsDouble();

List<MyObject> filteredElements = list
                .stream()             
                .filter(o-> o.getAverage() > average)
                .collect(Collectors.toList());