 |
 | Home | Contact Us | FAQ | Search & Site Map | Link to Us |
 | Sign In | Join | Other 45 Sites in Network |
Java Forum / First Aid / April 2005Cross tabulation - someone please help if you can
Hi, I need advise re the best data structure to use for representing data in aggregated form. I am looking to do something like this Year 1 Year 2 number percent number percent ... MAJ 66 0.0 61 0.0 ... GRAD 0 0.0 0 0.0 ... NOTA 0 0.0 3 0.0 ... TOTAL 66 100.0 66 100.0 ... The example shown above is my output from a current program. I will not include it here because I don't want bring too much focus to the program itself. The problem is that I can't get the percent values to come out correctly although the number part is right. I really don't know what else to do after two days of looking at this code. The question is, what data structure is best suited to do this type of work. I tried using a hashtable [] where each hash table in the array held data for one group. So for the data above I would have five hashtables in the array each holding Maj, Grad, Nota and Total data items as keys. The value part is a class that has number and percent along with some other details. Any help will be much appreciated. Please let me know if it is helpful for you to see the code so far. -Reuben > Hi, > [quoted text clipped - 7 lines] > NOTA 0 0.0 3 0.0 ... > TOTAL 66 100.0 66 100.0 ... Am I reading you correctly here? Because from what you're saying above I would figure something like: class YearStats { final int maj; final int grad; final int nota; YearStats(int maj, int grad, int nota) { this.maj = maj; this.grad = grad; this.nota = nota; } public int total() { return int+maj+grad; } public int asPercent(int value) { return (total() / 100) * value; } } Which you'd then keep in a simple List: ArrayList<YearStats> years = new ArrayList<YearStats>(); years.add(new YearStats(66,0,0)); years.add(new YearStats(61,0,3)); ... etc ... And then you could print 'em: for (YearStats year : years) { System.out.println("Year "+i+"); System.out.println("MAJ " + year.maj+" (" + year.asPercent(maj) + ")"); System.out.println("GRAD "+year.grad+" (" + year.asPercent(grad) + ")"); System.out.println("NOTA "+year.nota+" (" + year.asPercent(nota) + ")"); System.out.println("TOTAL "+year.total()); } Not sure if that's what you meant? > Any help will be much appreciated. Please let me know if it is helpful > for you to see the code so far. > > -Reuben Always :)  Signature[Ross A. Bamford] [ross AT the.website.domain] Roscopeco Open Tech ++ Open Source + Java + Apache + CMF http://www.roscopec0.f9.co.uk/ + info@the.website.domain Thanks! your thoughts are good but I think I failed to properly communicate the problem. Now I am forced to post the code and perhaps that will help see what i am trying to do. The sample table I provided is actually what the final output should look like except the percent calculation needs to be right. I am posting here the main method. Everything works well except the first row output in the percent column all resolve to 100% and while debuging, I have determined that most of the percent values are not even being innitialized. Where is the problem coming from? public static void main(String[] args) throws IOException{ //read an array of fall cohorts File inFile = new File(args[0]); ArrayList<FallCohorts> fallcohort = new ArrayList<FallCohorts>(); try{ BufferedReader in = new BufferedReader(new FileReader(inFile)); String line = null; while((line = in.readLine()) != null){ FallCohorts cohort = new FallCohorts(); cohort.readData(line); fallcohort.add(cohort); } in.close(); } catch(EOFException e){ e.printStackTrace(); } final String NOT_IN_ATTENDANCE = "NOTA"; final String GRADUATED = "GRAD"; final String ENROLLED_SAME_PROGRAM = "SMAJ"; final String ENROLLED_OTHER_PROGRAM = "OMAJ"; final String TOTAL = "TOTAL"; String [] rowNames = new String[5]; rowNames[0] = ENROLLED_SAME_PROGRAM; rowNames[1] = ENROLLED_OTHER_PROGRAM; rowNames[2] = GRADUATED; rowNames[3] = NOT_IN_ATTENDANCE; rowNames[4] = TOTAL; Hashtable [] h = new Hashtable[13]; //initialize array for(int i = 0; i < h.length; i++){ h[i] = new Hashtable(); h[i].put(ENROLLED_SAME_PROGRAM, new TabulateTerm()); h[i].put(ENROLLED_OTHER_PROGRAM, new TabulateTerm()); h[i].put(GRADUATED, new TabulateTerm()); h[i].put(NOT_IN_ATTENDANCE, new TabulateTerm()); h[i].put(TOTAL, new TabulateTerm()); //check keys are innitializing correctly // Enumeration e = h[i].keys(); //while(e.hasMoreElements()){ //System.out.println(i + " " + e.nextElement()); //} } //Loop and tabulate TabulateTerm row, columnTotal; for(FallCohorts f : fallcohort){ //column 1 if(f.getCohortStartYear() == f.getYearTerm()){ row = (TabulateTerm) h[0].get(f.getGroup()); columnTotal = (TabulateTerm) h[0].get(TOTAL); row.setTerm(f.getYearTerm()); row.setGroup(f.getGroup()); row.setNumber(row.getNumber() + 1); columnTotal.setNumber(columnTotal.getNumber() + 1); row.setPercent(row.getNumber()/columnTotal.getNumber()); columnTotal.setPercent(100); } //column 2 else if((f.getCohortStartYear() + 8) == f.getYearTerm()){ row = (TabulateTerm) h[1].get(f.getGroup()); columnTotal = (TabulateTerm) h[1].get(TOTAL); row.setTerm(f.getYearTerm()); row.setGroup(f.getGroup()); row.setNumber(row.getNumber()+1); columnTotal.setNumber(columnTotal.getNumber()+1); row.setPercent(row.getNumber()/columnTotal.getNumber()); columnTotal.setPercent(100); } //column 3 else if((f.getCohortStartYear()+ 10) == f.getYearTerm()){ row = (TabulateTerm) h[2].get(f.getGroup()); columnTotal = (TabulateTerm) h[2].get(TOTAL); row.setTerm(f.getYearTerm()); row.setGroup(f.getGroup()); row.setNumber(row.getNumber()+1); columnTotal.setNumber(columnTotal.getNumber()+1); row.setPercent(row.getNumber()/columnTotal.getNumber()); columnTotal.setPercent(100); } //column 4 else if((f.getCohortStartYear()+ 18) == f.getYearTerm()){ row = (TabulateTerm) h[3].get(f.getGroup()); columnTotal = (TabulateTerm) h[3].get(TOTAL); row.setTerm(f.getYearTerm()); row.setGroup(f.getGroup()); row.setNumber(row.getNumber()+1); columnTotal.setNumber(columnTotal.getNumber()+1); row.setPercent(row.getNumber()/columnTotal.getNumber()); columnTotal.setPercent(100); } //column 5 else if((f.getCohortStartYear()+ 20) == f.getYearTerm()){ row = (TabulateTerm) h[4].get(f.getGroup()); columnTotal = (TabulateTerm) h[4].get(TOTAL); row.setTerm(f.getYearTerm()); row.setGroup(f.getGroup()); row.setNumber(row.getNumber()+1); columnTotal.setNumber(columnTotal.getNumber()+1); row.setPercent(row.getNumber()/columnTotal.getNumber()); columnTotal.setPercent(100); } //column 6 else if((f.getCohortStartYear()+ 28) == f.getYearTerm()){ row = (TabulateTerm) h[5].get(f.getGroup()); columnTotal = (TabulateTerm) h[5].get(TOTAL); row.setTerm(f.getYearTerm()); row.setGroup(f.getGroup()); row.setNumber(row.getNumber()+1); columnTotal.setNumber(columnTotal.getNumber()+1); row.setPercent(row.getNumber()/columnTotal.getNumber()); columnTotal.setPercent(100); } //column 7 else if((f.getCohortStartYear()+ 30) == f.getYearTerm()){ row = (TabulateTerm) h[6].get(f.getGroup()); columnTotal = (TabulateTerm) h[6].get(TOTAL); row.setTerm(f.getYearTerm()); row.setGroup(f.getGroup()); row.setNumber(row.getNumber()+1); columnTotal.setNumber(columnTotal.getNumber()+1); row.setPercent(row.getNumber()/columnTotal.getNumber()); columnTotal.setPercent(100); } //column 8 else if((f.getCohortStartYear()+ 38) == f.getYearTerm()){ row = (TabulateTerm) h[7].get(f.getGroup()); columnTotal = (TabulateTerm) h[7].get(TOTAL); row.setTerm(f.getYearTerm()); row.setGroup(f.getGroup()); row.setNumber(row.getNumber()+1); columnTotal.setNumber(columnTotal.getNumber()+1); row.setPercent(row.getNumber()/columnTotal.getNumber()); columnTotal.setPercent(100); } //column 9 else if((f.getCohortStartYear()+ 40) == f.getYearTerm()){ row = (TabulateTerm) h[8].get(f.getGroup()); columnTotal = (TabulateTerm) h[8].get(TOTAL); row.setTerm(f.getYearTerm()); row.setGroup(f.getGroup()); row.setNumber(row.getNumber()+1); columnTotal.setNumber(columnTotal.getNumber()+1); row.setPercent(row.getNumber()/columnTotal.getNumber()); columnTotal.setPercent(100); } //column 10 else if((f.getCohortStartYear()+ 48) == f.getYearTerm()){ row = (TabulateTerm) h[9].get(f.getGroup()); columnTotal = (TabulateTerm) h[9].get(TOTAL); row.setTerm(f.getYearTerm()); row.setGroup(f.getGroup()); row.setNumber(row.getNumber()+1); columnTotal.setNumber(columnTotal.getNumber()+1); row.setPercent(row.getNumber()/columnTotal.getNumber()); columnTotal.setPercent(100); } //column 11 else if((f.getCohortStartYear()+ 50) == f.getYearTerm()){ row = (TabulateTerm) h[10].get(f.getGroup()); columnTotal = (TabulateTerm) h[10].get(TOTAL); row.setTerm(f.getYearTerm()); row.setGroup(f.getGroup()); row.setNumber(row.getNumber()+1); columnTotal.setNumber(columnTotal.getNumber()+1); row.setPercent(row.getNumber()/columnTotal.getNumber()); columnTotal.setPercent(100); } //column 12 else if((f.getCohortStartYear()+ 58) == f.getYearTerm()){ row = (TabulateTerm) h[11].get(f.getGroup()); columnTotal = (TabulateTerm) h[11].get(TOTAL); row.setTerm(f.getYearTerm()); row.setGroup(f.getGroup()); row.setNumber(row.getNumber()+1); columnTotal.setNumber(columnTotal.getNumber()+1); row.setPercent(row.getNumber()/columnTotal.getNumber()); columnTotal.setPercent(100); } //column 13 else if((f.getCohortStartYear()+ 60) == f.getYearTerm()){ row = (TabulateTerm) h[12].get(f.getGroup()); columnTotal = (TabulateTerm) h[12].get(TOTAL); row.setTerm(f.getYearTerm()); row.setGroup(f.getGroup()); row.setNumber(row.getNumber()+1); columnTotal.setNumber(columnTotal.getNumber()+1); row.setPercent(row.getNumber()/columnTotal.getNumber()); columnTotal.setPercent(100); } } //print heading row for(int i = 0; i < rowNames.length; i++){ System.out.print(rowNames[i] + "\t"); //print row data for(int j = 0; j < h.length; j++){ row = (TabulateTerm)h[j].get(rowNames[i]); System.out.print(row.toString()); } System.out.println(); } } > Thanks! > [quoted text clipped - 7 lines] > values are not even being innitialized. Where is the problem coming > from? Well, looking through that nothing jumps out immediately. It's a bit unconventional but I can't think of a reason it shouldn't work. Where to go next: TabulateTerm - is that your own class? It seems to me the most likely place to start. How/where is the percentage actually calculated? From what (variables)? Where do they get their values from? I assume you're using both the column and the row - are you sure they're the correct (paired) column and row? Check your code is executed: A simple System.out.println within the if/elseif construct(s) might help to trace things. Perhaps the values aren't initialized because none of the ifs match? Another idea is to println() your Hashtable.toString() once it's populated, to check it's laid out how you expect. Hope that helps, Ross Signature[Ross A. Bamford] [ross AT the.website.domain] Roscopeco Open Tech ++ Open Source + Java + Apache + CMF http://www.roscopec0.f9.co.uk/ + info@the.website.domain Well, unconventional? I have no doubt, but I am open to conventional suggestions of how I might do this. Thanks for your suggestion. It turns out, as it always does that I was creating my own woes. the setPercent method of the TabulateTerm class expects a double whereas I was passing the result of an int divided by an int. In most other languages that I know, c++ for example, you would get automatic conversion. The trouble with this operation in java is that no warnings or errors were being generated but instead it simply throws the result away - I think. I will have to spend some time to investigate exactly what it was doing. Thanks again, -Reuben > Well, unconventional? I have no doubt, but I am open to conventional > suggestions of how I might do this. Thanks for your suggestion. [quoted text clipped - 11 lines] > > -Reuben Glad you're closer to a solution :) And as for unconventional, whats the box for if you can't think outside it - big believer myself :) You shouldn't ever get a result thrown away without so much as a warning - Similar rules to c++ do apply in Java (although of course they differ in many details) but basically, primitive conversion is carried out explicitly where it doesn't involve loss of precision, but where an implicit conversion would result in such a loss, a compile-time error is generated. You might try inserting a couple of temporary S.o.println() calls, to sanity check the two parameters you are dividing before you divide, and then again inside the method when you first see the result as a double. I'd guess you'll find the original arguments are unset. Or of course set a watch if debugging be your pleasure. Cheers, Ross Signature[Ross A. Bamford] [ross AT the.website.domain] Roscopeco Open Tech ++ Open Source + Java + Apache + CMF http://www.roscopec0.f9.co.uk/ + info@the.website.domain Free MagazinesGet these publications absolutely FREE for up to 12 months. There are no hidden fees and no obligation. Simply choose a title, complete the application form and submit it. Read more ...
|
Reply to this Message
|
 Sign In
 Join
 My Latest Posts My Monitored Threads My Blog My Photo Gallery My Profile My Homepage Start New Thread Enable EMail Alerts Rate this Thread
|
©2008 Advenet LLC Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.