Bug? "Number of Hits in Book" graphs seem messed up
I've been mousing over the vertical bars in the "number of hits in book" graph, and the bar heights seem to be all messed up. A book with 25 occurrences will be the tallest bar, while a book with 32 occurrences is much shorter. Etc.
(unless I'm totally misunderstanding this graph - I'm assuming that the taller bars would represent more uses)
Comments
-
Could you post a picture? I ran a search and graphed it and it appeared ok to me.
Are you using the graph bible search feature?
0 -
Here's a simple example because it has so few uses (attached). The word occurs 6 times: 3x in Proverbs, 2x in Hosea, 1x in II Samuel. But the tallest bar is clearly Hosea. When I mouse over that tallest bar it gives the correct info "Hosea (2)" but the visualization (which is the point of the graph) is not helpful because it is incorrect.
I'll admit my ignorace - I don't know how to use the "graph bible search feature."
To admit more ignorance - it seems to me that the usage graph we are talking about only shows up some of the time in the word study and exegetical guides, not all of the time? Shouldn't it show up in the "word for word" section of the the exegetical guide?
0 -
TimothyLovegrove said:
he word occurs 6 times: 3x in Proverbs, 2x in Hosea, 1x in II Samuel. But the tallest bar is clearly Hosea
Hi Timothy.
We call this a "lemma density graph" (at least that's what I remember calling it in 3.x when we conceived of them there). The graphic style is called a "sparkline", a term coined by Edward Tufte to describe very small graphics that attempt to convey a large amount of information (see Tufte's web site for more info: http://www.edwardtufte.com/bboard/q-and-a-fetch-msg?msg_id=0001OR&topic_id=1 )
The height does not simply represent count, it has to do with the "density" of the term in each book. So while the word occurs 3x in Proverbs, the densest hits are in Hosea with 2 hits. With words that occur infrequently, sometimes the density graphs appear askew, but once one takes into account it isn't a simple hit graph but a density graph, things start to make more sense (like the shorter books having larger bars for the same number of words).
Hope it helps.
Rick Brannan
Data Wrangler, Faithlife
My books in print0 -
Aha! Now it makes sense, thank you!
Any idea why it seems to appear sometimes and not others?
0 -
Rick Brannan said:
Hope it helps.
Thanks for the explanation. I've wondered that myself and your explanation makes perfect sense.
0 -
TimothyLovegrove said:
I'll admit my ignorace - I don't know how to use the "graph bible search feature."
Not a problem! You'll probably find this feature helpful.
If you click on one of the "textual searches" at the bottom of the Bible Word Study (or if you run a Bible search on the word), you'll see a place that says "Graph results." You can click this (see screenshot) and it will bring up a graph of the word's usage. You can change the type of graph to get a density graph like Rick mentioned, or to get the number of hits in each book, and a variety of other uses.
0 -
TimothyLovegrove said:
Aha! Now it makes sense, thank you!
You're welcome. I also realized some further explanation is in this blog post: http://blog.logos.com/archives/2005/12/sometimes_its_t.html
TimothyLovegrove said:Any idea why it seems to appear sometimes and not others?
I'll leave that to the developers and internal testers who I know are monitoring these posts (y'all are awesome!).
Rick Brannan
Data Wrangler, Faithlife
My books in print0
