# R and 3D plots

Christensen (Talk | contribs) (→3D Plots) |
Christensen (Talk | contribs) (→3D Plots) |
||

Line 64: | Line 64: | ||

Use '''help(scatterplot3d)''' to get the full list of parameters and options for the function. |
Use '''help(scatterplot3d)''' to get the full list of parameters and options for the function. |
||

+ | |||

+ | |||

+ | === 3D and colors === |
||

+ | Replace the '''scatterplot3d''' parameters with the following in order to get (nicer) colorful 3D plots: |
||

+ | <pre> |
||

+ | scatterplot3d(xyz, highlight.3d=TRUE, col.axis="blue", col.grid="lightblue", main="scatterplot3d", pch=20) |
||

+ | </pre> |

## Latest revision as of 16:59, 22 November 2004

The R Project for Statistical Computing is a convenient tool for plotting experimental data. If you are using Debian, you can find the R core .deb package, otherwise you can download the R source code and compile it yourself.

## [edit] Basics

**Starting R**: Once R is installed you start it by typing R. This will give you an R prompt.

**Quitting R**: q()

**HTML help**: help.start()

**Searching online help**: help.search("what you don't know")

**Assignment**: variable <- value, eg. x <- 42

**Show content of a variable**: variable, eg. x

**Loading data from a file**: read.table("filename") (assuming that your data file contains data only, where each line describes one point in a multi-dimensional space).

**Loading data from a file into a variable**: xyz <- read.table("filename") (you knew that!)

## [edit] 3D Plots

In order to plot in 3D you need to add a package called *scatterplot3d* to R. You should download scatterplot3d package from one of the CRAN mirrors and leave it as a .tar.gz file. Then as root type the following from command-line:

# R CMD INSTALL scatterplot3D_version.tar.gz

Now, to use the package you start R and type:

> library(scatterplot3d)

In order to load the package. Hereafter, you can use the function scatterplot3d(). Assume that you have a file in the following format:

# an example file for 3D plots with R and scatterplot3d x1 y1 z1 x2 y2 z2 x3 y3 z3 .. ..

To do the plot the following sequence of commands will do the trick:

# R > library(scatterplot3d) > xyz <- read.table("examplefile.data") > scatterplot3d(xyz) > xyz x1 y1 z1 x2 y2 z2 x3 y3 z3

Assuming everything looks good and you want to save the plot:

> postscript("greatlookinggraph.eps") > scatterplot3d(xyz) > dev.off() > q()

The command **dev.off()** turns off the postscript device and flushes unwritten data. There are multiple devices, the default on UNIX is the x11 device, which produces nice windows. A device is started and set to the default graphical output device by simply typing the name of the device followed by brackets. Remember to issue a **dev.off()** when you are done with a device. A list of available devices is shown if you type **help(Devices)** -- lookout by the way R is **case-sensitive** hence, **help(devices)** will leave you with nothing but an error message.

Use **help(scatterplot3d)** to get the full list of parameters and options for the function.

### [edit] 3D and colors

Replace the **scatterplot3d** parameters with the following in order to get (nicer) colorful 3D plots:

scatterplot3d(xyz, highlight.3d=TRUE, col.axis="blue", col.grid="lightblue", main="scatterplot3d", pch=20)