How to calculate max concurrency at any point of time

Hi,

I am having  a file(sorted from old timestamp to latest) like below
512912.00             6/1/2010 11:20:02  1.00                  
512913.00             6/1/2010 11:20:03  1.00                  
512912.00             6/1/2010 11:20:03  2.00                
512914.00             6/1/2010 11:22:53  1.00                  
512917.00             6/1/2010 11:23:38  1.00                  
512918.00             6/1/2010 11:25:04  1.00                  
512919.00             6/1/2010 11:25:13  1.00                  
512920.00             6/1/2010 11:26:40  1.00                  
512921.00             6/1/2010 11:26:49  1.00                  
512953.00             6/1/2010 11:41:05  1.00                  
512954.00             6/1/2010 11:41:19  1.00                  
512943.00             6/1/2010 11:41:21  2.00                  
512861.00             6/1/2010 11:41:56  2.00                  
512958.00             6/1/2010 11:42:35  1.00                  
512959.00             6/1/2010 11:43:39  1.00                  
512895.00             6/1/2010 11:48:56  2.00                  
512906.00             6/1/2010 11:48:56  2.00                  
512901.00             6/1/2010 11:49:56  2.00        

where
1column unique userid's
2nd column timestamp
3rd column:
1.00 and 4.00 = logged in
2.00 and 3.00 =logged out

Please do let me know any script which will be able to calculate the maximum number of concurrency from the entries in file
login.zip
LVL 4
Ramakanta SahooAsked:
Who is Participating?

Improve company productivity with a Business Account.Sign Up

x
Advertise Here
 
ValeriConnect With a Mentor Commented:
This is the optimal solution with java:
1. make such as class:
public class Event {
    private long id;
    private Date evDate;
    private int type;
}
2. Create a Map
Map<Long, Event> events = new TreeMap<Long, Event>();
3. Assuming that for each event of type 2 you have event of type 1 before that, you have to do the following:
Loop throught the file and the following:
1) If you find event of type 1 add it to the "events", if you find event of type 2, remove the event of type 1 from the map and calculate the period using as start date the date from event type "1" and as end date the date from event type "2".
After having this period you have to disperse it to the the period you have already defined and divided by intervals by 1 hour let say. So if the user is logged in at 15:20 and logged out at 17:30, during the dispersion you have to increase by 1 the hours 15:00, 16:00 and 17:00.
After that you have to find the maximum users per hour. Ofcourse the dispersion coluld be not by hour but by every 30 min let say, or by 10 min...

0
 
ksivananthCommented:
in Java use scanner to read the file and check if the last token in each line is 1 or 4 if so, add them into concurrent!

http://www.java-tips.org/java-se-tips/java.util/scanning-text-with-java.util.scanner-3.html
0
 
Ramakanta SahooAuthor Commented:
Thanks. but if I am calculating max concurrent users what shall be the algorithm as users are also logging out too. so I can not blindly add all 1,4 to find maximum concurrent usage over a date range

It will be great if anyone can provide me a script or java or perl code to do it.

I found some thing where you can calculate in DB but dont know how to put it for this file.
SELECT MAX(ConcurrentUsers)
FROM  (SELECT     a.StartTime, COUNT(*) AS ConcurrentUsers
       FROM       MyTable AS a
       INNER JOIN MyTable AS b
             ON   b.StartTime <= a.StartTime
             AND  b.EndTime >= a.EndTime
       WHERE      a.StartTime >= @StartOfPeriod
       AND        a.StartTime <= @EndOfPeriod
       GROUP BY   a.StartTime) AS x

http://www.sqlnewsgroups.net/sqlserver/t839-calculate-concurrent-users-between-date-time-range.aspx
0
Free Tool: Subnet Calculator

The subnet calculator helps you design networks by taking an IP address and network mask and returning information such as network, broadcast address, and host range.

One of a set of tools we're offering as a way of saying thank you for being a part of the community.

Access It Now
 
ksivananthCommented:
got you, you can have a map( userid as key and count as value ) and then finally travers through the map and add up all the counts( non negative )
0
 
ksivananthConnect With a Mentor Commented:
on the other hand you can export this file to DB table( with delimeter ) and execute the the query!
0
 
Ramakanta SahooAuthor Commented:
Am new to this. It will be great if I can get the code to do that.
DB import is not an  option as it takes huge amount of time if the data is huge and its time taking too.
0
 
ksivananthCommented:
>>Am new to this. It will be great if I can get the code to do that.

why don't you try coding using the example of Scanner and come back if you struck?
0
 
Ramakanta SahooAuthor Commented:
Thanks I will try it but i need to prepare for java first. :) .  i wanted some script in perl which will be able to do it. The issue is I have very less time in my hand. But sure i will try it.

Any other help appreciated.! thanks
0
 
CEHJConnect With a Mentor Commented:
You could model this using nested set. The largest set would give you the largest concurrency
0
 
Ramakanta SahooAuthor Commented:
Wanted Code as this was urgent requirement. Now learning and coding took my time. Any way thanks for the fantastic ideas.
0
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.