Hello,
I am new to power automate, and have been trying to build a model that can recognize information in a table from an image and converting it to text (one of the criteria however is that information in a table must map to each other). I'm encountering a seemingly basic problem - that is applying a value in a subgroup to multiple groups. I am trying to apply the last column's first value to the first three items, and the second value to the last three items but the grid cuts the data so that item 1, 3, 4, 6 do not have a corresponding tube ID. How can I fix this given that in advanced tagging I cannot double tag a single value? The grid also does not allow me to delete certain boundaries.
Here is the preview:
Alternatively if there is another way to map item 1, 3, 4, 6 (Part number) to Tube ID using another method please let me know as well (hoping for as much automation as possible)! Thanks!
Solved! Go to Solution.
Hi @KylieRen - welcome to Power Automate and thank you for sharing this use case!
Indeed, today it's not possible to tag multiple times the same cell. Are there always 6 rows in the table you want to extract? If so, when you read the results in a Power Automate cloud flow, you can assume the value for Tube ID for cells 1 & 3 is the same as cell 2.
If number of rows can vary, something you can try is to tag the tube ID in the cell for the first row it corresponds to. If the AI model is able to learn it this way, then when you process each row in a cloud flow you can assume the Tube ID cell value is equal to the previous one until a different, non-empty value is returned.
Let us know if you try these and what are the results.
Hi @KylieRen - welcome to Power Automate and thank you for sharing this use case!
Indeed, today it's not possible to tag multiple times the same cell. Are there always 6 rows in the table you want to extract? If so, when you read the results in a Power Automate cloud flow, you can assume the value for Tube ID for cells 1 & 3 is the same as cell 2.
If number of rows can vary, something you can try is to tag the tube ID in the cell for the first row it corresponds to. If the AI model is able to learn it this way, then when you process each row in a cloud flow you can assume the Tube ID cell value is equal to the previous one until a different, non-empty value is returned.
Let us know if you try these and what are the results.