add exponent 2 to percent of variance explained equation

vignettes/01_04_rmarkdown.Rmd

@@ -47,7 +47,7 @@ knit_hooks$set(setPch = function(before, options, envir) {
 opts_chunk$set(setPch = TRUE)
 ```
 
-We will again hide this in future documents, but this shows how to set a default figure width and height, as well as setting plot margians. 
+We will again hide this in future documents, but this shows how to set a default figure width and height, as well as setting plot margins. 
 
 
 ```{r global-plot,warning=FALSE, message=FALSE}

vignettes/01_10_exploratory-analysis.Rmd

@@ -53,11 +53,11 @@ biocLite("alyssafrazee/RSkittleBrewer")
 
 * Use plots as often as possible
 * Use tables for phenotype data
-* Look for 
-  * Missing values
-  * Outlier values
-  * Mislabeled samples
-  * Naming consistency
+* Look for: 
+    * Missing values
+    * Outlier values
+    * Mislabeled samples
+    * Naming consistency
 
 ### Make the plots pretty
 
@@ -142,15 +142,15 @@ dim(edata)
 
 Here we see that there are a lot of outliers
 ```{r boxplot}
-boxplot(log2(edata+1),col=2,range=0)
+boxplot(log2(edata+1),col=2,range=0, las=2, cex.axis=0.8)
 ```
 
 We can also look at this sample by sample with histograms
 
 ```{r histograms}
 par(mfrow=c(1,2))
-hist(log2(edata[,1]+1),col=2)
-hist(log2(edata[,2]+1),col=2)
+hist(log2(edata[,1]+1),col=2, cex.main=0.8)
+hist(log2(edata[,2]+1),col=2, cex.main=0.8)
 ```
 
 Or with density plots
@@ -180,7 +180,7 @@ We can remove rows that are mostly zero and notice any differences in the distri
 ```{r}
 edata = as.data.frame(edata)
 filt_edata = filter(edata,rowMeans(edata) > 1)
-boxplot(as.matrix(log2(filt_edata+1)),col=2)
+boxplot(as.matrix(log2(filt_edata+1)),col=2, las=2, cex.axis=0.8)
 ```
 
 ### Check for obvious data mixups

vignettes/02_03_dimension-reduction.Rmd

@@ -74,7 +74,7 @@ names(svd1)
 
 ## Look at the percent variance explained
 
-The percent of variance explained is given by $\frac{d_{ii}}{\sum_{j}d_{jj}^2}$
+The percent of variance explained is given by $\frac{d_{ii}^2}{\sum_{j}d_{jj}^2}$
 ```{r}
 plot(svd1$d,ylab="Singular value",col=2)
 plot(svd1$d^2/sum(svd1$d^2),ylab="Percent Variance Explained",col=2)