Running Kideval only after spa codes

[Sebastian Gajardo]

Hello, we have transcripts of child and mother interactions. We were able to run mor on those transcripts. Now our question is what is the mlu words, freq types, and TD words after certain communicative acts which are coded with a %spa dependent tier. We are trying to separate $SDS:* from all other spa codes.  What I have been doing is running a kwal to separate the codes (SDS vs all other spa codes). But the issue is I cannot run kideval to get mlu words, freq types, and TD words because the output kwal file is not in proper chat format. So before I am able to run kideval I have to fix all the files to be in legal chat form. Is there an easier way to run kideval (or any other commands that will give the variables) for certain dependent tiers?

[Brian MacWhinney]

Sebastian,

If you add the +d switch to your KWAL command it should output legal CHAT format.

-- Brian MacWhinney
Teresa Heinz Professor of Cognitive Psychology,
Computational Linguistics,
and Modern Languages, CMU

> On Feb 26, 2020, at 1:48 PM, Sebastian Gajardo <seb.ga...@gmail.com> wrote:
>
> Hello, we have transcripts of child and mother interactions. We were able to run mor on those transcripts. Now our question is what is the mlu words, freq types, and TD words after certain communicative acts which are coded with a %spa dependent tier. We are trying to separate $SDS:* from all other spa codes. What I have been doing is running a kwal to separate the codes (SDS vs all other spa codes). But the issue is I cannot run kideval to get mlu words, freq types, and TD words because the output kwal file is not in proper chat format. So before I am able to run kideval I have to fix all the files to be in legal chat form. Is there an easier way to run kideval (or any other commands that will give the variables) for certain dependent tiers?
>

> --
> 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/0e6749ea-7d2e-4f01-812c-7cabf3e64699%40googlegroups.com.
> <127bn.cha>

[Sebastian Gajardo]

but it won't allow me to run a kideval because all the files are missing an all the header tiers (@Begin, @Language ect.)

On Wednesday, February 26, 2020 at 1:22:54 PM UTC-6, macw wrote:

Sebastian,

    If you add the +d switch to your KWAL command it should output legal CHAT format.

-- Brian MacWhinney
Teresa Heinz Professor of Cognitive Psychology,
Computational Linguistics,
and Modern Languages, CMU

> On Feb 26, 2020, at 1:48 PM, Sebastian Gajardo <seb.ga...@gmail.com> wrote:
>
> Hello, we have transcripts of child and mother interactions. We were able to run mor on those transcripts. Now our question is what is the mlu words, freq types, and TD words after certain communicative acts which are coded with a %spa dependent tier. We are trying to separate $SDS:* from all other spa codes.  What I have been doing is running a kwal to separate the codes (SDS vs all other spa codes). But the issue is I cannot run kideval to get mlu words, freq types, and TD words because the output kwal file is not in proper chat format. So before I am able to run kideval I have to fix all the files to be in legal chat form. Is there an easier way to run kideval (or any other commands that will give the variables) for certain dependent tiers?
>
> --
> 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 chib...@googlegroups.com.

[Sebastian Gajardo]

The kwal I am using are:

For SDS: kwal +t*chi +d +t%spa *.kwal.cex +re +s$SDS:* +t%mor +t%gra +f

For all other spa codes: kwal +t*chi +d +t%spa *.kwal.kwal.cex +re +s$* +t%mor +t%gra +f +...@NewGrantNonSDS.cut (cut file attached)

Then from those output files I have to add header tiers on all files to then run kideval

On Wednesday, February 26, 2020 at 1:22:54 PM UTC-6, macw wrote:

Sebastian,

    If you add the +d switch to your KWAL command it should output legal CHAT format.

-- Brian MacWhinney
Teresa Heinz Professor of Cognitive Psychology,
Computational Linguistics,
and Modern Languages, CMU

> On Feb 26, 2020, at 1:48 PM, Sebastian Gajardo <seb.ga...@gmail.com> wrote:
>
> Hello, we have transcripts of child and mother interactions. We were able to run mor on those transcripts. Now our question is what is the mlu words, freq types, and TD words after certain communicative acts which are coded with a %spa dependent tier. We are trying to separate $SDS:* from all other spa codes.  What I have been doing is running a kwal to separate the codes (SDS vs all other spa codes). But the issue is I cannot run kideval to get mlu words, freq types, and TD words because the output kwal file is not in proper chat format. So before I am able to run kideval I have to fix all the files to be in legal chat form. Is there an easier way to run kideval (or any other commands that will give the variables) for certain dependent tiers?
>
> --
> 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 chib...@googlegroups.com.

[Brian MacWhinney]

Sebastian,
Yes, I see the problem. Ideally the +d switch should copy over the headers, but it only copies over @Begin and @End. Even adding +t@* doesn’t help. I think Leonid will have to look into this.

-- Brian MacWhinney
Teresa Heinz Professor of Cognitive Psychology,
Computational Linguistics,
and Modern Languages, CMU

> 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/ff478967-2b82-42da-865a-67a719c69fcd%40googlegroups.com.
> <NewGrantNonSDS.cut>

[Sebastian Gajardo]

I ran kideval from the kwal out file after redownloading clan and it works now without having to add the header tiers. Sorry for the trouble

[Leonid Spektor]

Sebastian,

You are using KWAL for filter tiers based on keyword(s). The +t option in KWAL is used to specify which tiers you want to look at for those keyword(s). To add addition tiers to the output that you do not want to be included in the search for keywords function you have to use +o option. For example your first command should be:

kwal +t*chi +d +t%spa *.kwal.cex +re +s$SDS:* +o@ +o%mor +o%gra +f

Leonid.

On Feb 26, 2020, at 14:40, Sebastian Gajardo <seb.ga...@gmail.com> wrote:

The kwal I am using are:

For SDS: kwal +t*chi +d +t%spa *.kwal.cex +re +s$SDS:* +t%mor +t%gra +f

For all other spa codes: kwal +t*chi +d +t%spa *.kwal.kwal.cex +re +s$* +t%mor +t%gra +f +s...@NewGrantNonSDS.cut (cut file attached)

Then from those output files I have to add header tiers on all files to then run kideval

On Wednesday, February 26, 2020 at 1:22:54 PM UTC-6, macw wrote:

Sebastian,

    If you add the +d switch to your KWAL command it should output legal CHAT format.

-- Brian MacWhinney
Teresa Heinz Professor of Cognitive Psychology,
Computational Linguistics,
and Modern Languages, CMU

> On Feb 26, 2020, at 1:48 PM, Sebastian Gajardo <seb.ga...@gmail.com> wrote:
>
> Hello, we have transcripts of child and mother interactions. We were able to run mor on those transcripts. Now our question is what is the mlu words, freq types, and TD words after certain communicative acts which are coded with a %spa dependent tier. We are trying to separate $SDS:* from all other spa codes.  What I have been doing is running a kwal to separate the codes (SDS vs all other spa codes). But the issue is I cannot run kideval to get mlu words, freq types, and TD words because the output kwal file is not in proper chat format. So before I am able to run kideval I have to fix all the files to be in legal chat form. Is there an easier way to run kideval (or any other commands that will give the variables) for certain dependent tiers?
>
> --
> 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 chib...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/chibolts/0e6749ea-7d2e-4f01-812c-7cabf3e64699%40googlegroups.com.
> <127bn.cha>

--
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/ff478967-2b82-42da-865a-67a719c69fcd%40googlegroups.com.