TM1 Forum

Hi all,

I created a TI to main a Product dimension. I just realised a serious bug.

For example, I have 2 products in a source CSV file.

1st row: Model Code = "ABC-123P", Dept = "XXX"
2nd row: Model Code = "ABC-123 P", Dept = "XXX"

Hence, the difference between 1st row and 2nd is the presence of space in Row 2. I'm concatenating Model Code with an underscore, follow by Dept to derive the Element Code. E.g. ABC-123P_XXX.

After updating the dimension, I noticed that product in 2nd Row is always created, and product in Row 1 is always missing. The element code shown is "ABC-123 P_XXX" i.e. with a space in between.

I altered the source file and reordered the two rows. Now, product "ABC-123P_XXX" will always be created instead as it is now in 2nd row.

Why can't TI treat the two model codes as uniquely different items? Is there some fixes or some ASCII function to resolve this issue? Please help!

Many thanks in advance!

Hi Harry,

As has been explained many many times previously on this forum, and probably to you once or twice as well; TM1 is both caSe and sp ac e insensitive.

Thus as far as TM1 is concerned
ABC-123P
and
ABC-123 P
and for that matter
a b c-123p
are all identical for the purposes of identifying a dimension element.

First check your source system - should these keys really be different? If so then you should consider replacing the spaces with another character such an an underscore (either in the source system when extracting, or in TI prior to creating the element.)

If this is indeed space insensitive, then why is it that the element code can be shown as "ABC-123 P_XXX"?

You can show the space, but TM1 will still see it as spaceless when it comes to using it. If you try and add the elements "no space" and "nospace" to a dimension manually you will get a duplicate element error, as you are attempting to save the same element twice as fas as TM1 is concerned.

I see. Thank you both for your quick reply!