Markdown document
=================
This is a text-document, with extension ".Rmd" --- R-markdown. 

The point is that the R-code is between lines starting with a line with three single back-quotes followed by {r} and ending with a line of 3 singe back-quotes. When this type of file is processed in R-studio by pressing "Knit HTML" or simply Shift-Ctrl-H, all the R-code will be run, and both the code and the results collected in a final HTML document.

# Here is a section

First we set some relevant prerequisites --- the opts_chunk$set is used to specify whether we want to keep the code - and we indeed do. Also remember to explicitly set the working directory and let R confirm it:
```{r}
opts_chunk$set(keep.source=TRUE)
library( Epi )
# setwd("c:/bendix/undervis/SPE/Intro")
```

## Here is a subsection

For documentation we list the folder where it all takes place and the list and version-names(!) of the attached packages, so we know if some of the calculations were done by package versions that later showed to contain errors:
```{r}
getwd()
print( sessionInfo(), l=F )
```

### Subsub-section

This is as far down as it goes.

# Subject matter

Then we can start the **substance** of the document to show what goes on:

The data are pulled out of thin air:
```{r}
xx <- rnorm(9638)
xx[1:35]
```
Note that we have generated `r length(xx)` random numbers from a normal distribution, and `r round(mean(xx>1)*100,1)`% of them are larger than 1.

Note that the listing is changed if you change the width-parameter - this is what you want to customize to fit with your usual printing style:
```{r}
options("width")
options( width=120 )
options("width")
xx[1:35]
options( width=77 )
```
Now we can make a simple plot:
```{r}
hist( xx, breaks=100, col=gray(0.3), border=gray(0.3) )
```

Note that the plot is automatically included in the html document. And that you should leave a blank line after the chunk with a plot.

We can make more than one plot, and control the height and width of the plot inside the curly brackets. **Then** we can make the two plots side-by-side:
```{r, fig.width=8, fig.height=4}
par( mfrow=c(1,2), mar=c(3,3,1,1), mgp=c(3,1,0)/1.5 )
hist(         xx  , breaks=100, col=gray(0.4), border=gray(0.4), main="" )
hist( log(abs(xx)), breaks=100, col=gray(0.2), border=gray(0.2), main="" )
```

At the end of the document you are writing, it is convenient to have a number of empty code-brackets stashed up, because it is a pain to write all those back-quotes in a Danish keyboard.

Also it is convenient to insert inside long stretches of code where you want to include some comments.
```{r}
```
```{r}
```
```{r}
```
```{r}
```
```{r}
```
```{r}
```