I think this could be an easy one…count and count distinct in UNIQUE
Regards,
Miguel
I need to send the results to a file, with other related metrics in UNIQUE. For example, for client (unique) I need to collect sales sum, number of invoices (count needed here), and so on…
little more complex … maybe the pivot gives you already what you need and the rest is not needed
distinct.transform (7.7 KB)
‘Count unique’ is on the wishlist for Unique. We could possibly add ‘Count distinct’ as well. I will make a note.
2 Likes
There are additional Unique options in the latest snapshot. Please try them and let us know if you have any issues:
Edit Fixed incorrect link.
I don’t understand the new options, can you kindly explain.
I created following input table:
using unique:
Output is like a pivot with count
If I try now other options I don’t get what’s going on:
It shows all the time the same.
You need to press the Apply button to get the option applied to the table.
Unique isn’t very useful unless you have at least 2 columns.
If you have a dataset like this:
| Key | Value1 |
|---|---|
| 001 | A |
| 001 | A |
| 001 | B |
| 002 | C |
| 002 | D |
And want to count how many values in Value1 are unique for each *Key value:
So:
- Key=A has 1 value that occurs once (B).
- key=B has 2 values that occurs once (C & D).
If you click Add count column, it will add an extra column telling you how many rows in the input dataset make up each key column.
Ps/ Keep Unique is not a particular good name. It should perhaps be Key column. And the Unique transform might be better called Aggregate or Group By. However we don’t really want to do major renaming, except perhaps during a major release.
1 Like
Thanks for explanation, now I get it. I suggest to add the explanation with the 2 columns minimum and so on to the documentation. In a few weeks I will be lost again.
