05-02-2011 09:45 AM
05-03-2011 12:35 AM
07-06-2011 11:12 AM
In EGuide, it requires multiple tasks. First you use the rank task to assign the sequential numbers with name (you may have to fiddle with the options so it won't use tied ranks). Then you do the concatenation in a Query Buider using the advanced expression builder.
07-09-2011 08:44 AM
You can also:
1. Using query builder, select the 'Name' variable and click 'select distinct values only' to create a new dataset.
2. User query builder and add a computed column 'count' defined with the function 'monotonic()'
3. Use query builder to merge back with the orginal dataset, and concatenate "Phone" and count.
This means you dont have to change the phone number to numeric.
07-11-2011 09:30 AM
Thank you Nick. I created the computed column, Count1, with the function, monotonic(t1.Name). The Count1 column assigned sequential numbers, but they were for an entire data set as shown below. In other words, it didn't start from 1 for Mary Adams.
Name Phone Count1
John Smith 111-222-3333 1
John Smith 123-525-8888 2
John Smith 222-444-9999 3
Mary Adams 333-888-7777 4
Mary Adams 337-841-1111 5
07-11-2011 05:16 PM
The trick is to first create a datasets with distinct values for the 'Name' variable, then apply the monotonic function and merge back to the original dataset. This means having two nodes in EG, but I cant think of a way to do it in one step.
Hope thats clearer.
07-14-2011 10:48 AM
What you can do is, starting from the outcome of the table that you have with created with monotonic, the following:
Determine the minimum value of count1 per name. In the results that you displayed that would be 1 for John Smith and 4 for Mary Adams.
Merge this info with the original outcome and compute a column that consists of the difference between count1, the minimum value and add 1.
This would result in:
Name Count1 Min_Count1 SeqNumber
John Smith 1 1 1
John Smith 2 1 2
John Smith 3 1 3
Mary Adams 4 4 1
Mary Adams 5 4 2
Hope this helps.