Filtering a column by number of characters or numeric

I have a column in my dataset with manufacturing lot numbers. I need to decode the lot numbers into things like shift, line, month etc. We changed the way we generate lot numbers a few years back, and the decoding is different, so I have to separate them out. I was thinking of filtering by number of characters, but I may be able to use the numeric portion to filter as well. One lot number is only numeric, whereas the other is alphanumeric. All ideas welcome. Thanks!

Old Lot Number (7 chars, alphanumeric): 11Z4216

New Lot Number (11 chars, numeric): 1612234131

Datasets

Connectors

Accepted answers

Thableaus

@mattcoleman

Fortunately, there's a very useful tool called RegEX that can ease your job.

Once you have the pattern, you can tell the tool what your parsing output should be.

I did it for both (Old and New) Lot Numbers. See solution:

Old Lot Number RegEX (Regular Expression):

(\d{2})([A-Z])(\d)(\d)(\d{2})

2 numbers, a letter, one number, one number, 2 numbers. That's what you're telling the tool to parse.

Then you define the column names in the configuration window, and it's done.

New Lot Number RegEX:

(\d{2})(\d{2})(\d{2})(\d)(\d)(\d)(\d)

2 numbers, 2 numbers, 2 numbers, one number, one number, one number, one number

Same thing as Old Lot Number.

Workflow appended.

Does this work well for you?

Cheers,

RegEX Parsing.yxmd

All comments

Thableaus

Hi @mattcoleman

It'd be better if you specified what your desired output is.

It seems to me you want to parse this Lot Number to a lot of columns (shift, line, month).

What are the rules? Could you share us an example?

Cheers,

mattcoleman

Yes, @Thableaus that's exactly what I'm trying to do. Maybe separating them out isn't necessary, but they are mixed together in a column called "Lot Number".