Logos 4.5b Beta 2 Morph Search Grid Results Puzzling ? ?

Follow-up to thread => grammatical search of the genitive case in the Greek New Testatment

Found a bit of a puzzle: changing from Morph Search

@N[^G] BEFORE 2 WORDS @NG

to Bible Search allows for Grid search, whose results have some oddities among Logos Greek Morphology resources:

image

Logos 4.5b Beta 2 on Mac and Windows appear to have identical oddities (adventure perhaps ?).

Looking at Matthew 1:1 in LDGNT, puzzled by Grid search result having a blank box.  Appears some Names have variant tagging (e.g. genitive is a bit inconsistent).

Keep Smiling [:)]

Comments

  • Tonya J Ross
    Tonya J Ross Member Posts: 1,490

    I'm probably just missing it, but is the puzzling behavior you're referring to the shift from 'before 2 words' to 'before 0-2 words'?  If so, are you still getting this incorrect behavior?  I can't reproduce that issue in 4.5b Beta 4.

  •  If so, are you still getting this incorrect behavior?

    Using Logos 4.5b Beta 4 on Mac and PC, have some puzzling search results.  For example, Matthew 1:1 is not found in LDGNT:

    image

    Puzzled by Matthew 1:17 not being found in Lexham Syntactic Greek New Testament.  Personally learned Names can have XP tagging (non-declinable), which can vary a bit between Greek resources.  For Matthew 4:4 was a bit surprised by results found in UBS4:

    image

    Also noted several more results that were not found, which appear to match search condition.

    Keep Smiling [:)]

  • Tonya J Ross
    Tonya J Ross Member Posts: 1,490

    Thank you for the screenshots.  I'll make sure this issue gets reported.

  • Bradley Grainger (Logos)
    Bradley Grainger (Logos) Administrator, Logos Employee Posts: 11,955

    Using Logos 4.5b Beta 4 on Mac and PC, have some puzzling search results.  For example, Matthew 1:1 is not found in LDGNT:

    image

    There appears to be an indexing error in the LDGNT: the word Βίβλος is not indexed. (Right-click it, choose Search this resource, and it will find 0 hits.) I've filed a bug for this case.

    Puzzled by Matthew 1:17 not being found in Lexham Syntactic Greek New Testament.


    For the LSGNT, no hit is found because "Prepositional" and "Phrase" count as intervening words. (This is by design; to search by word proximity in Greek, search a plain Greek NT.)

    For Matthew 4:4 was a bit surprised by results found in UBS4:

    image

    RE: UBS4 "Before two words ?" -- this looks like an indexing bug; possibly the same as in LDGNT.

    RE: LSGNT: SBL "Puzzling not found in Morph Search results" -- there are intervening words.

    RE: LDGNT: "Not Found ?"  -- the second morph is NDSM, which doesn't match the query.

    RE: Lexham SGNT and OpenText GNT "Not Found ?" -- the intervening characters count as words for the query.

     

  • There appears to be an indexing error in the LDGNT: the word Βίβλος is not indexed. (Right-click it, choose Search this resource, and it will find 0 hits.) I've filed a bug for this case.

    Thanks [8-|] for confirming a bug.

    Puzzled by Matthew 1:17 not being found in Lexham Syntactic Greek New Testament.

    For the LSGNT, no hit is found because "Prepositional" and "Phrase" count as intervening words. (This is by design; to search by word proximity in Greek, search a plain Greek NT.)

    Appreciate puzzle explanation [8-|]

    Keep Smiling [:)]

  • Bradley Grainger (Logos)
    Bradley Grainger (Logos) Administrator, Logos Employee Posts: 11,955

    RE: UBS4 "Before two words ?" -- this looks like an indexing bug; possibly the same as in LDGNT.

    This indexing bug will be fixed in 4.5c Beta 3. There are a handful of instances in UBS4 Int. where word proximity queries will give incorrect results (as shown above); to fix it you will need to run "rebuild index" in 4.5c Beta 3.

    (LDGNT turned out to be a resource error, not an indexing bug; it will require a resource update to fix.)

  • Keep Smiling 4 Jesus :)
    Keep Smiling 4 Jesus :) MVP Posts: 23,113

    RE: UBS4 "Before two words ?" -- this looks like an indexing bug; possibly the same as in LDGNT.

    This indexing bug will be fixed in 4.5c Beta 3. There are a handful of instances in UBS4 Int. where word proximity queries will give incorrect results (as shown above); to fix it you will need to run "rebuild index" in 4.5c Beta 3.

    (LDGNT turned out to be a resource error, not an indexing bug; it will require a resource update to fix.)

    Looking forward to rebuilding indexes in 4.5c Beta 3 on Mac and PC while wondering how many hours for each ? (currently anticipating PC to finish indexing hours sooner than Mac)

    Thanks for LDGNT insights; looking forward to resource update soon (using Logos historical word usage).

    Keep Smiling [:)]

This discussion has been closed.