Skip to main content

How to avoid @isini false positives

Hi,
I'm using @isini to find the word "bar", however, it is bringing so many false positives (.i.e. "Barbara") that I would like to avoid. I remember having heard some tricks to solve this in one of the webinars, but I don't remember how and did not find my notes. Any suggestion?  

Brian Element Thu, 11/11/2021 - 09:41

Hi Zeus,

Are you looking for an exact match?  If so you can use the == instead of the @isini as the == will give you an exact match.  If you are not worried about case you should probably change both sides to upper or lower case.  Also if you are looking for items like bar, bars, etc you could use the @len function to only compare with words that are four or five characters.  Can you explain more what you are looking for so that I can give you better suggestions.  Is this an exact match or what type of matches are you looking for with bar?

Zeus Thu, 11/11/2021 - 17:46

In reply to by Brian Element

Hi Brian,
To ilustrate this better, using the following table, I need to identify only lines 1, 4 and 5. Right now, using @isini, I get all of them, I would like to excluse the false positives.
        Description
1       bar
2      Trip to Santa Barbara
3      Soundbar project
4      Took customer to bar
5      Hotel bar
6      Gift - Barbie doll
 
Thank you

Brian Element Thu, 11/11/2021 - 19:28

In reply to by Zeus

Hi Zeus, it looks like you should be using the search function.  It is in the data tab.  You enter the word and select the field to search in.  Here is an example using your file:

This is the results which I believe is what you are looking for:

Zeus Tue, 11/16/2021 - 15:03

In reply to by Brian Element

Hi Brian,
Thank you for your reply. However, I need other options different to extraction; I have done some workaround using @isini ("bar ", " bar", etc.). I remember having seen in a webinar what I looking for, but unfortunately I did not find my notes. I think it was a combination of @len, or @left, or something similar.
 
Thank you for your help
Zeus

Brian Element Mon, 11/22/2021 - 05:28

Hi Zeus,

So what other options do you need other then extracting those values?  The @len and @left work for the entire field so they won't work in your instance.  If there was only one word then you could use the @len but that is not the case.  You can alwasy add a @recno to your file, do the above extract and then join them back to the original file based on the @recno.