Create a large number of fields in a datatype by upload?

I need to create 930 fields in a datatype, and I don’t want to do this manually. Any suggestions?

Yes. Don’t. That is a jerky answer, but I can’t imagine the use case where having an object/type with that many fields makes sense. Can you give a little more insight as to why you think that is the right solution?

Anyway, to answer your question for real, I know of no way to automate the creation of fields themselves.

Marc

1 Like

Well, I do have a legitimate use case for this. Each field represents a numerical value for a particular record. Each value is unique, so i m not aware of a better way of doing this.

So, for each record how many of those 930 fields will contain data?

on average, about 750

Are they name value pairs ?

So does each numerical value have a descriptor of what it is ?

I would be trying to do this as a list field to be honest.

1 Like

Without knowing more, making suggestions are like stabbing in the dark, but I’ll close my eyes and try anyway. If you are only going to be reading and updating the fields (and not searching or filtering) you could find logical ways to group them (say 100 at a time) and store them as delimited strings and use Regex to get and set values.

This would allow you to only have to create a field for each group vs. each data point.

But presuming there is no more reasonable way to store the data and you have to create a table with 930 fields, you could use a Browser Plugin scripting tool to manipulate the Data Tab in bubble.

Lookup field name and type
Click Create New Field
Enter field name in input
Pick type from Dropdown
Click Create Button
Wait 2 seconds
Loop

Marc

Yes, each numerical value has a descriptor. Think of it like this. The data, originally, is a table. Columns B - AW (48 total), and rows 2 - 24 (23 total) for a total of 1,104 unique combinations. There are 176 always invalid combinations, for a total of 928 possible unique numerical values per record. And there are a couple of hundred records so far.

Since they are in a XY table format, I made the labels (fields) correspond to the columns and rows in Excel, so the 1st field is B2, and so on…for 928 fields…plus a couple of more to identify the record by code and name.

I hope that adds some clarity.

If I grouped them by column, there would be 48 groups/data types. I don’t think this would simplify anything.

Part of the problem is my limited knowledge in this area, so I am certain that I am not explaining this properly. It’s a problem of each record having its own 2x2 table. This led me to flatten out the table, putting the values in one row, hence the need for 928 fields. Perhaps I am thinking about this all wrong.

How does one go from 2x2 to 3x3 in a database conceptually?

Thank you @NigelG. I’ll look into the list field. That would potentially make things much easier…48 list fields.

In testing this, I have run into the bracket problem, which apparently still has not been addressed…

This is enormously frustrating. Can’t get the data to upload into the database using a common csv.

Seriously, this shouldn’t be this difficult.

It doesn’t work with lists of texts (for example) but you can quite easily upload multiple tables and link them together.

You if you upload table a and table b, you can then a bulk API to link a and b.

will it work with lists of numbers?

I have no idea how to bulk API.

I am having trouble visualising your data, this 2x2 table.

Can you give an example or two ?

And even if I did, I don’t see how this would work.

Apparently, we are not allowed to share csv files here?