PDA

View Full Version : Root list separated by homonym?



anotherOTguy
03-29-2013, 11:53 PM
How can I get a list of all the Hebrew verbal roots that occur with a particular parsing, listing the number of occurrences for each root, BUT separated by homonym?

I get everything except the homonym separation by doing the following:

.*@vp* (to get all piel verbs, for example)
Then in the "words" tab --> "wildcard expansion of command line", I get all of the verbal roots, with the number of occurrences, but it combines homonyms.

Is there any automated way to get the exact same list, but separated by homonym? So in the Piel example, instead of DBR having only one entry with 1092 hits, there would be three entries DBR-1 (to drive away) with 6 hits, DBR-2 (to speak) with 1085 hits, andf DBR-3 (to have descendants) with 1 hit. And so on for every root that occurs with that parsing.

I can do that manually, but it requires running separate searches for each homonym of each root that shows up in the initial search.

Todah!
John

anotherOTguy
04-03-2013, 01:42 PM
I figured out a way to do it.



.*@vp*+*Ha*
lists all of the homonym 1 roots that occur in the Piel in "words" tab --> "wildcard expansion of command line"
.*@vp*+*Hb*
the same, but for homonym 2
.*@vp*+*Hc*
the same, but for homonym 3
.*@vp*+*Hd*
the same, but for homonym 4
.*@vp*
The tricky part is that any root that does not have a homonym code will show up in none of searches 1-4. Instead, it will only show up in .*@vp* (which will also get all of the others). So one would search .*@vp*, and then import the resulting list of roots (words tab --> wildcard expansion of command line) into the word list, and then subtract out all of the roots that were obtained in the homonym searches (steps 1-4) to get the roots with no homonym code.

Jim Wert
04-04-2013, 03:03 PM
I figured out a way to do it.



.*@vp*+*Ha*
lists all of the homonym 1 roots that occur in the Piel in "words" tab --> "wildcard expansion of command line"
.*@vp*+*Hb*
the same, but for homonym 2
.*@vp*+*Hc*
the same, but for homonym 3
.*@vp*+*Hd*
the same, but for homonym 4
.*@vp*
The tricky part is that any root that does not have a homonym code will show up in none of searches 1-4. Instead, it will only show up in .*@vp* (which will also get all of the others). So one would search .*@vp*, and then import the resulting list of roots (words tab --> wildcard expansion of command line) into the word list, and then subtract out all of the roots that were obtained in the homonym searches (steps 1-4) to get the roots with no homonym code.


For number 5 above, this seems to work for me:
.*@vp*+*Hx*
To validate it I ran this little test:


Query
Forms
Hits


.*@vp*+*Ha*
363
621


.*@vp*+*Hb*
324
1723


.*@vp*+*Hc*
23
27


.*@vp*+*Hd*
5
6


.*@vp*+*Hx*
1830
4113







.*@vp*
2545
6490


If your calculator agrees with mine, the total forms and hits found by the first 5 queries sum to the same number of forms and hits found by the last (.*@vp*) query.

I hope the above table displays OK; I made a number of attempts to format it for this forum, and went with what displayed best on my preview.