Hi all,
I'm trying to find an optimal tool to try and Bin a number of records into say 5 or 10 group of equal records. I feel the Tile tool is only able to bin into 3 groups based on the standard deviation.
For example say I have a dataset that is items sold:
RecordID | Items Sold |
1 | 10 |
2 | 5 |
3 | 7 |
4 | 4 |
5 | 5 |
6 | 1 |
7 | 8 |
8 | 9 |
9 | 7 |
10 | 9 |
11 | 8 |
12 | 7 |
13 | 7 |
14 | 2 |
15 | 2 |
16 | 7 |
17 | 3 |
18 | 5 |
19 | 6 |
20 | 1 |
21 | 3 |
I want to find out what range I should set to say have 3 groups with a equal count in each group (in this case 7 records in each group). Is there a certain tool that would be able to auto-calculate that?
Thanks!
Hello @whitkrieng ,
Have you tried the tile tool with the Equal Records Method?
I got three groups with 7 records each:
Gabriel
Thanks for your response, I think the Equal Records clusters it randomly. So in my instance, I want an algorithmic way to group the "Items Sold" in 3 buckets determined by the distribution of the "Items Sold". If say there are 7 records for 1,2, and 3 # of "Items Sold" then I would construct a bucket of "1-3 Items Sold". Then the next bucket could "4-5 Items Sold" and last "6-10 Items" Sold.
Hi @whitkrieng ,
you may be talking about the multi-field binning tool. I've built an example around a 7 group distribution based on Items sold:
I've attached the build and you can mess around with the configuration of the binning tool.
Hope this helps,
M.