You’re offline. This is a read only version of the page.
11
Hello,
I need to create a csv table which has a specific format / template with a 3 row header, for each column (like shown above).
| StartDate | Status | Call_SID | Interaction_ID |
| Start Date | Response Type | Call_SID | Interaction_ID |
| {"ImportId":"startDate","timeZone":"Europe/London"} | {"ImportId":"status"} | {"ImportId":"Call_SID"} | {"ImportId":"Interaction_ID"} |
| «values» | «values» | «values» | «values» |
This is a specific format we use to import legacy data.
The first three rows are considered header for my platform. As the values are only expected in the 4th row of the file.
I have managed to create the csv all right with the right data in the value cells. All except for the header that I can't manage to have in 3 different rows / cells (all I could manage was to have all strings in the same cell).
Can you please advise on a way to achieve this?
Hello @sandracosta
I did a test on my side that it is working:
In this example, we have an excel file with 2 columns (Region, Country). The steps are:
1) Define headers (3 rows)
2) List Excel Rows
3) Select the Region and Country columns
4) Create a CSV table
5) Replace "Region,Country" string by the headers.
Hope it helps!
Ferran
Thanks so much for your fast replies.
I tried to go in the direction you appointed but I guess I am not as proficient in Power Automate as I should have been to solve this flow. My flow only got to this:
This flow stops in the “csv table” creation with the error “The property 'columns' must be specified unless the 'from' property value is an array of objects.”. However, in the last compose output I don’t seem able to identify the columns and hence can’t specify them.
The output from the union / input to the csv table comes out like this:
[ [ "StartDate", "EndDate", "Q1", "Q2" ], [ "Start Date", "End Date", "De 0 a 10 indique pf como recomendaria a Galp aos seus amigos e familiares, com base na sua experiência com este serviço de atendimento (em que 0 é não recomendaria e 10 é recomendaria de certeza)", "De 1 a 5 indique pf o seu grau de satisfação com o atendimento que acabou de ter? (em que 1 é Muito Insatisfeito e 5 é Muito Satisfeito)" ], [ "{\"ImportId\":\"startDate\",\"timeZone\":\"Europe/London\"}", "{\"ImportId\":\"endDate\",\"timeZone\":\"Europe/London\"}", "{\"ImportId\":\"QID1\"}", "{\"ImportId\":\"QID2\"}" ], { "Start": "44161.4674189815", "End": "44161.4674189815", "Q1": "9", "Q2": "5" }, { "Start": "44161.7289467593", "End": "44161.7289467593", "Q1": "9", "Q2": "5" }, { "Start": "44161.4358796296", "End": "44161.4358796296", "Q1": "5", "Q2": "2" }, { "Start": "44161.3903703704", "End": "44161.3903703704", "Q1": "9", "Q2": "5" }, … |
A problem that I had was that I wasn’t able to initiate an array type variable. I kept getting the error “The variable 'ExtraHeader' of type 'Array' cannot be initialized or updated with value of type 'String'. The variable 'ExtraHeader' only supports values of types 'Array'.” with this script.
To solve this, I composed the extra header with createArray function. And “union” it to the select of the original file. But not sure if its the best practice.
You could do something like this:
Explanation:
Create a new array that is the same format as the array that is going to be made into CSV.
Add your additional header records into that array.
Union that array with your existing data and then put the result into the "Create CSV Table" Step.
Hello @sandracosta
Before creating the csv table, you could add 2 new rows at the top of the table, with header data in rows 2 and 3, using, for example, a compose action or appending to an array one (if you share your flow and the format of your data, we might provide more detailed information).
Hope it helps!
Ferran
#1
Michael E. Gernaey
497
Super User 2025 Season 1
#2
David_MA
436
Super User 2025 Season 1
#3
Riyaz_riz11
244
Super User 2025 Season 1
Share
Report
All responses (
Answers (
