00_01_3_IntroToR_practical.qmd

---
title: "P01. Introduction to R, part 3"
---

**A. Read in the same data as before**

This file can be downloaded from [here](data/TB_stats.txt).

```{r eval = F}
myTBdata <- read.table("TB_stats.txt", header=TRUE)
```

**B. Plot the mortality in HIV negative against HIV positive** check the plot function help file

```{r eval = F}
?plot
```

`plot` is a generic function, and depending on what type of data you pass the function R will use different sub-functions (you dont need to worry about how it handles this!).

```{r eval = F}
# make the plot
plot(x=myTBdata$HIV_neg_TB_mortality, y=myTBdata$HIV_pos_TB_mortality)
```

**C. Add meaningful axes labels**

```{r eval = F}
plot(x=myTBdata$HIV_neg_TB_mortality, 
	 y=myTBdata$HIV_pos_TB_mortality,
	 xlab="Mortality in HIV negative people",
	 ylab="Mortality in HIV positive people")
```

**D. Add a meaningful title**

```{r eval = F}
plot(x=myTBdata$HIV_neg_TB_mortality, 
	 y=myTBdata$HIV_pos_TB_mortality,
	 xlab="Mortality in HIV negative people",
	 ylab="Mortality in HIV positive people",
	 main="Comparison of mortality in HIV negative and positive")
```

**E. Change the colour of the points to red**

```{r eval = F}
plot(x=myTBdata$HIV_neg_TB_mortality, 
	 y=myTBdata$HIV_pos_TB_mortality,
	 xlab="Mortality in HIV negative people",
	 ylab="Mortality in HIV positive people",
	 main="Comparison of mortality in HIV negative and positive",
	 col="red")
```

**F. It's hard to see the numbers because some are small and some very large** Using a log scale is useful for that.You can either log the values and re-plot, or use the log option in plot()

```{r eval = F}
plot(x=myTBdata$HIV_neg_TB_mortality, 
	 y=myTBdata$HIV_pos_TB_mortality,
	 xlab="Mortality in HIV negative people",
	 ylab="Mortality in HIV positive people",
	 main="Comparison of mortality in HIV negative and positive",
	 col="red",
	 log="xy")
```

**G. Now let's make a different kind of plot** Show the distribution of Total_TB_mortality in a histogram and then change the x axis label. Note, same options as before.

```{r eval = F}
hist(myTBdata$Total_TB_mortality)

hist(myTBdata$Total_TB_mortality,
	 xlab="Number")
```

Now add a meaningful title to the plot

```{r eval = F}
# Add your code here
```

**H. Check what other aspects of the histogram you can change**

```{r eval = F}
?hist
```

Then change the color to "blue" in the last plot

```{r eval = F}
hist(myTBdata$Total_TB_mortality,
	 xlab="Number",
	 main="Total TB mortality",
	 col="blue")
```

**I. Now let's plot a histogram of mortality per 1000**

**Hint:** calculate it as in the previous practical

```{r eval = F}
# Add your code here
```

And add a title, and x axis label

```{r eval = F}
# Add your code here
```

change the color to something different hint: to find more colours, run "colors()" or google "Colors in R"

```{r eval = F}
# Add your code here
```

**J. Now let's show both histograms at the same time** you need to make a call to "par", short for parameters, setting the plot parameter "mfrow" (Multi-Figure ROW-wise) gives 1 row, and 2 columns of plot

```{r eval = F}
par(mfrow=c(1,2))
hist(myTBdata$Total_TB_mortality,
	 xlab="Number",
	 main="Total TB mortality",
	 col="blue")
```

cut and paste your plot code from I. here and run it. Then resize the plot window and see what happens

**K. Export the figure and save it as a PNG with a useful name**

**Hint:** use the Export button in the plot window

**L. R has functions for every kind of plot** for example:

```{r eval = F}
?barplot
?boxplot
?contour
```

and stackoverflow.com has a lot of comments and help on every kind of plot

**Advanced plotting exercises**\
Make a plot where:

-   x=Total_TB_mortality and y=HIV_pos_TB_mortality

-   both aces are on the on the log scale

-   colour the points

-   add axis labels and a title

```{r eval = F}
# Add your code here
```

add HIV_neg_TB_mortality on the same y axis, in a different colour.

**Hint:** use points(). see `?points` for information

```{r eval = F}
# Add your code here
```

do you need to change the y axis label? i.e. does it still make sense now that it shows negative and positive mortality?

Answer:

Some of the points no longer fit on the graph. Why is this? You need to alter the y limit (ylim), which is an option of plot. What value will you choose?

**Hint:** the maximum value that the data go to change the ylim of the plot.

```{r eval = F}
# Add your code here
```

The plot now has 2 data sets in different colours, so it needs a legend check the help of legend (there's a lot of options!)

**Hint:** use x="topright" instead of setting the x and y values for location.

**Hint:** use the option "fill" to change the colours

```{r eval = F}
# Add your code here
```

The solutions can be accessed [here](00_01_3_IntroToR_solutions.qmd).