Day 2

Author: psolymos

README.md

@@ -33,7 +33,8 @@ Each day will consist of 3 sessions, roughly one hour each, with short breaks in
 | **Day 2** |  **Behavioral complexities**  |    |
 |  |  1. Statistical assumptions and nuisance variables  |  [Slides](https://peter.solymos.org/qpad-workshop/day2-1-intro.pdf)  |
 |  |  2. Behavioral complexities  |  [Notes](https://peter.solymos.org/qpad-workshop/day2-2-behavior.html)  |
-|  |  3. Removal modeling techniques  |  [Notes](https://peter.solymos.org/qpad-workshop/day2-3-removel.html)  |
+|  |  3. Removal modeling techniques 1.  |  [Notes](https://peter.solymos.org/qpad-workshop/day2-3-removal.html)  |
+|  |  3. Removal modeling techniques 2.  |  [Notes](https://peter.solymos.org/qpad-workshop/day2-3-mixtures.html)  |
 | **Day 3** |  **The detection process** |    |
 |  |  1. Distance sampling |    |
 |  |  2. Calibrating population density  |    |

day2-1-intro.Rmd

@@ -283,15 +283,15 @@ And a lot more that we'll cover later
 
 # Why do we need simulation
 
-Probabilistic simulation is useful to test how well a stat method is working _if_ the assumptions are met.
+Statistical techniques are expected to provide unbiased estimates when the data generation process follows the assumptions of the model. __Testing this is standard practice for any new technique.__
 
-It is not quite as useful for assessing how robust the method is when the assumptions are _not_ met.
+But testing if the __model assumptions are realistic__ is rarely evaluated with the same rigor.
 
-You need to link to biological mechanisms to do sensitivity analysis.
+For that we need to link to biological mechanisms to do sensitivity analysis. This is a lot more complex than a coin toss.
 
 ***
 
-# It is a good pass time
+# Capturing complexity
 
 \centering
 _"I've yet to see any problem, however complicated, which when you looked at it the right way didn’t become still more complicated."_ --- Poul Anderson, Call Me Joe
@@ -309,6 +309,16 @@ _"Love the simulation we’re dreaming in"_ --- Dua Lipa, Physical
 
 ***
 
+# Some history
+
+Movement trajectories of bird flocks to mitigate mortality near airports and wind farms.
+
+Beth Rigby ([2016](https://conservancy.umn.edu/handle/11299/193439)): _Simulating Effects of Imperfect Detectability in Bird Surveys_
+
+But not a lot more...
+
+***
+
 # bSims goals
 
 - Allow easy __testing of statistical assumptions__ and explore effects of violating these assumptions

day2-3-removal.Rmd

@@ -211,7 +211,7 @@ Singing rates of birds vary with time of day, time of year, breeding status, and
 
 Let's fit a couple of time-varying models using `DAY` and `TSSR` as covariates:
 
-```{r beh-Me,cache=TRUE}
+```{r beh-Me}
 Me1 <- cmulti(Y | D ~ DAY, X, type="rem")
 Me2 <- cmulti(Y | D ~ TSSR, X, type="rem")
 ```

day2-4-mixtures.Rmd

@@ -10,7 +10,7 @@ load("data/josm-data.rda")  # JOSM data
 set.seed(1)
 ```
 
-```{r eval=FALSE}
+```{r eval=TRUE}
 ## data analysis from day2-3 file
 yall <- Xtab(~ SiteID + Dur + SpeciesID, 
   josm$counts[josm$counts$DetectType1 != "V",])
@@ -385,8 +385,9 @@ xall <- bsims_all(
   density=c(1, 1, 0),
   vocal_rate=phi,
   move_rate=1, movement=0.2,
-  tau=tau,
-  tint=tbr, rint=rbr)
+  tau=0.8,
+  tint=c(3,5,10), 
+  rint=c(0.5, 1, 1.5))
 xall
 ```