diff --git a/R_statistics_2022/R_stats_script.R b/R_statistics_2022/R_stats_script.R
index 219944b..08cac83 100644
--- a/R_statistics_2022/R_stats_script.R
+++ b/R_statistics_2022/R_stats_script.R
@@ -868,7 +868,7 @@ plot(
 # To use spiec-easi for network inference, we will need to prepare the data into a specific R object type.
 # Also, since it uses clr-transformation to deal with compositionality, we need to use the filtered data set
 # that was already used in the RDA
-physeqo_pro <-phyloseq(
+physeqo_pro <- phyloseq(
   otu_table(otu_pro_filt, taxa_are_rows = T),
   tax_table(as.matrix(tax_pro[rownames(otu_pro_filt), ]))
 )
@@ -1001,7 +1001,7 @@ cluster.pick.rep <- lapply(
 names(cluster.pick.rep) <- paste0("M", 1:length(cluster.pick.rep))
 
 # Look e.g. at taxonomic composition
-i = 2
+i = 1
 View(tax_pro[cluster.pick.rep[[i]], ])
 
 # Correlation of eigengenes with environmental parameters
@@ -1038,9 +1038,9 @@ plot(
     to = c(1, 10)
   ),
   vertex.label = NA, 
-  vertex.color = ifelse(node_stats$degree > 15, "red", "blue"),
+  # vertex.color = ifelse(node_stats$degree > 15, "red", "blue"),
   # vertex.color = ifelse(node_stats$closeness > 0.045, "red", "blue"),
-  # vertex.color = ifelse(node_stats$betweenness > 5000, "red", "blue"),
+  vertex.color = ifelse(node_stats$betweenness > 5000, "red", "blue"),
   edge.color = "grey", 
   edge.width = abs(E(adj.g.pos)$weight) * 5,
   main = ""
diff --git a/R_statistics_2022/R_stats_slides.pdf b/R_statistics_2022/R_stats_slides.pdf
index 88c1ed5..fbb953b 100644
Binary files a/R_statistics_2022/R_stats_slides.pdf and b/R_statistics_2022/R_stats_slides.pdf differ
diff --git a/R_statistics_2022/machine_learning_slides.pdf b/R_statistics_2022/machine_learning_slides.pdf
new file mode 100644
index 0000000..4b7fc8f
Binary files /dev/null and b/R_statistics_2022/machine_learning_slides.pdf differ