Hi,
What's now:
I have formula that validate data from 4 columns. If string in all 4 columns match, it marks it as a true otherwise as a false. That 4 columns comes from 4 different files.
What I want:
Since not everytime all 4 files will be loaded (instead i load an empty template file with just column names and no data) I need to modify that so it will be processing only those columns which were loaded. And i don't want to create a superlong formula were i manually code all cases.
Do you have any ideas that could help?
Thanks in advance
Try to use the multi row formula, select only the string fields
Hi @I2ufu5
Here is how you can achieve it.
Input: 4 columns with null where data is not loaded
Workflow:
1. Generate record ID as a key
2. Use transpose tool to covert 4 column to rows and set all other columns as keys
3. Use summarize tool, config countdistinctnonnull of values (which will check only non null values). If the count distinct is 1 then all values are same/match.
4. Use formula tool to calculate flag. True if countdistinct is 1 and false if not.
5. Using join tool to map all other columns back.
Hope this helps 🙂
That look nice but i need a distinction between null value and column(file) not loaded at all.
What i mean by that is when I load files number 1,2,3 and not 4, workflow will produce that output:
Record | File1 | File2 | File3 | File4 | Verification |
1 | A | A | A | null | True |
2 | A | null | A | null | False |
First row will be true because File4 wasn't loaded (no data in whole column)
Second row will be false because File2 was loaded but it had a null value at that index.
Hi @I2ufu5
I have modified the workflow to include the scenario.
Workflow:
1. I have added a data cleanse before transpose tool to remove null columns
2. Changed summarize tool to count distinct including null. Now it been also considered as variation.
Hope this helps 🙂
User | Count |
---|---|
19 | |
15 | |
13 | |
9 | |
8 |