Count combinations of codes in the same tier

[Melania S. Masià]

Hi everyone,

I have coded the use of referential expressions in a %ref tier. This tier includes seven codes with different values. When I use FREQ to extract the data, I get counts for each code and value per file, as expected (that is, a row for each file in the CSV file). However, I need the counts for the combinations of codes. I have seen in this conversation that wildcards can be used to generate counts of combinations, but that seems rather complicated with seven different codes. Is there a way to make each line of the %ref tier an observation and extract that data independently of the rest of the %ref lines in the file?

I've attached a minimal example and the desired CSV output in case it makes clearer what I mean. Thanks in advance!

Melania

[Leonid Spektor]

Hi Melania,

From your example output it looks like you want each %ref tier to be treated as one file. So, each %ref tier gets one row in the output. Currently, as you have noticed, the whole file gets a row in the output.

At this time FREQ can't do what you want. But, it might be possible to change FREQ to do what you want. I will let you know if it can be done.

Are you using Mac or Windows PC computer?

Leonid.

--
You received this message because you are subscribed to the Google Groups "chibolts" group.
To unsubscribe from this group and stop receiving emails from it, send an email to chibolts+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/chibolts/7a05605c-7044-40f4-a197-252dc70aed3an%40googlegroups.com.