====== An example for the jitter effect ======

Adding a small noise to the original data might be helpful in visualizing data, especially for categorical data. The scatterplot of two categorical variables just m x n points, from which we do not know how many sample points are actually in these graphical symbols. If we ''jitter()'' the data before drawing a scatterplot, the effect can be better.

===== Animation =====

<ani je http://i288.photobucket.com/albums/ll181/xieyihui/jitter_effect/ png 30 480 0.5|Jitter Effect for Categorical Data>
Some small random numbers are added to the original data (which only take values on ''{1, 2, 3}''). If we do not jitter the data, the scatterplot will merely consist of nine points. As the process of jittering is random, this animation has shown its 30 possible results.
</ani>

===== R code =====

Refer to the example introduced in [[start#define_custom_functions|the start page of R package ''animation'']].