I'm struggling to describe this problem so it's proving elusive to search for a solution. I trust, however, that there's a way forward with Alteryx. Here's the scenario:
We have layouts that are effective for periods of time. When new ones arrive, they succeed prior versions of those layouts. The data has a layout ID which is unique to each event, and a prior layout ID along with effective dates. The effective dates are not necessarily reliable are are included here for illustration purposes.
I need to link these sequences of layouts together as a 'layout group' so I can run analysis on what has changed with these layouts. I've been playing with a multi-row formula but am struggling to put my thumb on the process to make this work.
Here's what initial data might look like:
Current Layout | Prior Layout | From | To |
14 | 12 | 12/1/2018 | 1/1/2019 |
13 | 11 | 12/1/2018 | 1/1/2019 |
12 | 10 | 11/1/2018 | 12/1/2018 |
11 | 9 | 11/1/2018 | 12/1/2018 |
10 | 8 | 10/1/2018 | 11/1/2018 |
9 | 7 | 10/1/2018 | 11/1/2018 |
8 | 6 | 9/1/2018 | 10/1/2018 |
7 | 5 | 9/1/2018 | 10/1/2018 |
6 | 4 | 8/1/2018 | 9/1/2018 |
5 | 3 | 8/1/2018 | 9/1/2018 |
4 | 2 | 7/1/2018 | 8/1/2018 |
3 | 1 | 7/1/2018 | 8/1/2018 |
And this is what I'm looking to transform it to, with the layout grouping to flag:
Current Layout | Prior Layout | From | To | Layout Group |
13 | 11 | 12/1/2018 | 1/1/2019 | 13 |
11 | 9 | 11/1/2018 | 12/1/2018 | 13 |
9 | 7 | 10/1/2018 | 11/1/2018 | 13 |
7 | 5 | 9/1/2018 | 10/1/2018 | 13 |
5 | 3 | 8/1/2018 | 9/1/2018 | 13 |
3 | 1 | 7/1/2018 | 8/1/2018 | 13 |
14 | 12 | 12/1/2018 | 1/1/2019 | 14 |
12 | 10 | 11/1/2018 | 12/1/2018 | 14 |
10 | 8 | 10/1/2018 | 11/1/2018 | 14 |
8 | 6 | 9/1/2018 | 10/1/2018 | 14 |
6 | 4 | 8/1/2018 | 9/1/2018 | 14 |
4 | 2 | 7/1/2018 | 8/1/2018 | 14 |
And as a side question, what in the world would this exercise be called? I'm sure it's been done before but am struggling to name it.
Solved! Go to Solution.
Hi @Sequencing
Make Group tool will make your job easy.
Using Make Group tool you separate the layout groups. Then you get the max value of each group and join that to your dataset.
Cheers,
That is absolutely perfect. Thank you!