Ted Lynch (tedlynch) wrote in casefile, @ 2018-05-07 10:47:00 |
|
|||
Entry tags: | character: theodora lynch, narration: network, player: sc |
Trying to run event duration/history modeling on a population with multiple spells, multiple failures, and a slowly decreasing time of exposure. Non-parametric, which means I have to run Cox, but of course the R is ten times more complicated for the Cox models. I think I'm running it wrong, and I think my populations are too small to find notable significance, which is unrealistic for any lone lab because of course I don't have 500 cows that were successfully impregnated via the new embryonic process, the failure rate is so high.
I've been up since yesterday, and there is not enough coffee in the world.
How is your Monday going, Fall City?
Related, I need 2 to 3 people to help me with a larger feeding experiment, I'll pay hourly and you'll get to hold a baby cow. Hit me up here if you are interested.