ehmatthes / sitka_irg_analysis

An analysis of the correlation between the Indian River stream gauge and landslide activity in the Sitka area.
4 stars 1 forks source link

Find a 100% threshold? #20

Closed ehmatthes closed 3 years ago

ehmatthes commented 3 years ago

I would guess there is a set of factors that would flag only true positives. It would miss some as well, but every event associated with those factors would also be associated with known slides.

This would allow two levels of criticality - one that lets you know when a slide is likely, and one that says every time conditions have been like this a slide has occurred.

I would guess this has more to do with total rise than with rate of rise.

To find this, look for 0 false positives.

ehmatthes commented 3 years ago

I can already see from the matrix of tries from 2.5 to 3.0 ft by 0.125 for rise, and 0.5 to 1.0 with .125 for rate, that the only sets with 0 false positives only found 1 true positive. This is not worth identifying.