diff --git a/07_final_assignment/.RData b/07_final_assignment/.RData
index 019e1f3..8d596fd 100644
--- a/07_final_assignment/.RData
+++ b/07_final_assignment/.RData
Binary files differ
diff --git a/07_final_assignment/baboonSimulation.R b/07_final_assignment/baboonSimulation.R
index b415679..ed528ff 100644
--- a/07_final_assignment/baboonSimulation.R
+++ b/07_final_assignment/baboonSimulation.R
@@ -5,6 +5,8 @@
 library(parallel)
 library(doParallel)
 library(foreach)
+library(itsadug)
+library(mgcv)
 enableJIT(3)
 
 
@@ -459,17 +461,11 @@
 #...RUN, MONKEY, RUN!
 cl <- makeCluster(num.cores)
 registerDoParallel(cl)
-
 foreach(i=1:num.cores) %dopar% do.experiment(i)
 
-##new version:
-#...RUN, MONKEY, RUN!
-#registerDoParallel(cores=num.cores)
-#foreach(i=1:num.cores) %dopar% do.experiment(i)
-
 
 ########################################################################################
-####                            VISUALIZATION & PLOTTING
+####                            HARVESTING DATA & VISUALIZATION I
 ########################################################################################
 
 harvest.data <- function(write.to.file=FALSE, folder="results_backup") {
@@ -498,6 +494,8 @@
   if(!any(names(data)==value)) {
     stop("'value' needs to be part of the data.frame")
   }
+  v.index = which(names(data)==value)
+  
   sequence <- round(sequence, 3)
   
   z <- matrix(nrow=length(sequence), ncol=length(sequence))
@@ -506,7 +504,6 @@
     
     alpha <- round(data[i, ]$alpha, 3)
     beta  <- round(data[i, ]$beta, 3)
-    value <- data[i, ]$value
 
     a.index <- get.index(sequence, alpha)
     b.index <- get.index(sequence, beta)
@@ -517,9 +514,8 @@
       print(beta)
       stop("sequence must contain alpha and beta values")
     }
-
-    z[a.index, b.index] <- data[i, ]$NumWordsLearned
-    z[b.index, a.index] <- data[i, ]$NumWordsLearned
+    z[a.index, b.index] <- data[i, v.index]
+    z[b.index, a.index] <- data[i, v.index]
   }
   return(z)
 }
@@ -527,7 +523,6 @@
 data = harvest.data(TRUE)
 
 
-
 mat1 = matrix(rexp(15*15, rate=.1), ncol=15, nrow=15)
 
 conditions <- c("GeneralAccuracy", "WordAccuracy", "NonwordAccuracy")
@@ -544,8 +539,65 @@
              col = topo.colors(100))
 }
 
+
 ########################################################################################
-####                            DATA ANALYSIS
+####                            MODEL FITTING & VISUALIZATION II
+########################################################################################
+
+### GeneralAccuracy
+m1.g = gam(GeneralAccuracy ~ s(alpha) + s(beta), data=data)
+summary(m1)
+
+m2.g = gam(GeneralAccuracy ~ s(alpha) + s(beta) + ti(alpha, beta), data=data)
+summary(m2.g)
+
+compareML(m1.g, m2.g)
+#m1 preferred
+
+vis.gam(m1.g, plot.type="contour", color="topo", main="GeneralAccuracy")
+
+
+
+### WordAccuracy
+m1.w = gam(WordAccuracy ~ s(alpha) + s(beta), data=data)
+summary(m1.w)
+
+m2.w = gam(WordAccuracy ~ s(alpha) + s(beta) + ti(alpha, beta), data=data)
+summary(m2.w)
+
+compareML(m1.w, m2.w)
+#m1 preferred
+
+vis.gam(m1.w, plot.type="contour", color="topo", main="WordAccuracy")
+
+
+### NonwordAccuracy
+m1.n = gam(NonwordAccuracy ~ s(alpha) + s(beta), data=data)
+summary(m1.n)
+
+m2.n = gam(NonwordAccuracy ~ s(alpha) + s(beta) + ti(alpha, beta), data=data)
+summary(m2.n)
+
+compareML(m1.n, m2.n)
+#m1 preferred
+
+vis.gam(m1.n, plot.type="contour", color="topo", main="NonwordAccuracy")
+
+
+### model criticism
+qqnorm(m1.g$residuals)
+qqnorm(m1.w$residuals)
+qqnorm(m1.n$residuals)
+# mostly ok
+
+acf(resid(m1.g))
+acf(resid(m1.w))
+acf(resid(m1.n))
+# ok
+
+
+########################################################################################
+####                            OTHER ANALYSIS SNIPPETS
 ########################################################################################
 
 
@@ -600,36 +652,5 @@
 anova(m2, m3)
 
 
-########################################################################################
-####                            IGNORE THIS BLOCK -> unused code
-########################################################################################
-
-
-
-res.matrix = matrix(nrow=length(a.seq), ncol=length(b.seq))
-counter = 1
-counter.a = 1
-for( a in a.seq) {
-  counter.b = 1
-  for (b in b.seq) {
-    res.matrix[counter.a, counter.b] = as.numeric(resultdat[counter, ]$GeneralAccuracy)
-    counter = counter + 1
-    counter.b = counter.b + 1
-  }
-  counter.a = counter.a + 1
-}
-
-View(res.matrix)
-
-# TODO: Durchlaufen lassen fuer ca 50k Trials fuer verschiedene alpha, beta zwischen 0 und 1
-# TODO: Plotten der verschiedenen Ergebnisse (num,acc) als heatmap 2d plot.
-# TODO: Plotten der Unterschiede zu den Vorgabeaffen und optimale alpha beta fuer diese finden.
-
-mat1 = matrix(rexp(15*15, rate=.1), ncol=15, nrow=15)
-#mat2 = matrix(rexp(200, rate=.1), ncol=50, nrow=50)
-
-image.plot(mat1, main="Word Accuracy", xlab="alpha", ylab="beta",
-           useRaster=TRUE, col = topo.colors(100))
-
 
 # EOF