Strange plot???

[mech3132]

Hi,

I'm playing around with the data to try to get a better visual handle of the data and found this weird effect (figure attached)...?

The dataset filtered by:

• Only samples that tested for Borellia sp. are included.
• number positive is raw number of positive tests for Borrellia sp.
• proportion positive is proportion of positive tests out of total tests (positive + negative; not NAs).
• I logged both axes, but did NOT add 1, so all zeros are removed here.

The weird diagonal line is found across years, domains, months, nlcdClass... can anyone think of what is causing this pattern? I'd like to know what causes it in case it's a systemic sampling problem... you can recreate this plot using the "TickMerged_hurdle.R", line 134/135 if you want to.

corr_propPositive_nPositive.pdf