essayez ce site scatter plot equivalent to pairs() by ggplot.
http://stackoverflow.com/questions/3735286/pairs-equivalent-in-ggplot2
Hadley recommends using the GGally package instead. It has a function, ggpairs that is a vastly improved pairs plot (lets you use non-continuous variables in your data frames). It plots different plots in each square, depending on the variable types:
Probability density function in R
http://www.cyclismo.org/tutorial/R/probability.html
Chi square goodness-of-fit test in R
http://ww2.coastal.edu/kingw/statistics/R-tutorials/goodness.html
Chi square independence test in R
http://ww2.coastal.edu/kingw/statistics/R-tutorials/independ.html
ANOVA in R
http://personality-project.org/r/r.anova.html
http://jeromyanglim.blogspot.ca/2010/05/abbreviations-of-r-commands-explained.html
250+ R commands.
Often times R users suffer from difficulties finding what commands are available for a specific process.
This lists 250+ R commands in table.
R color palettes
rainbow()
topo.colors()
terrain.colors()
heat.colors()
my.palette <- terrain.colors(100)
display.brewer.all()
my.palette <- brewer.pal(n, name) For example, my.palette <- brewer.pal(9, "Blues") However, the number of colors is limited up to 9, in most colors. Then use this.
more.color.palette <- colorRampPalette(brewer.pal(9,"Blues"))(100)
http://stackoverflow.com/questions/3832850/color-schemes-in-r
http://www.compbiome.com/2010/12/r-using-rcolorbrewer-to-colour-your.html
RColorBrewer palette names.
Change the label of legend not by changing the data.
# data
grp <- gl(n=4,k=20,labels=c("group a","group b","group c", "group d"))
value <- runif(n=80, min=10, max=150)
outcome <- cut(value,2)
data <- data.frame(grp,value,outcome)
# Option 1
# breaks should be exactly the same as the levels of the factor
ggplot(data, aes(grp, fill=outcome)) + geom_bar() +xlab("group") +
ylab("number of subjects") +
scale_fill_discrete("Serologic response",
breaks=c("(12.1,79.7]","(79.7,147]"), # should be the same as the levels of the factor
labels=c("double negative", "positive for a and/or b")
)
# Option 2
# simpler because factor() takes care of the levels of the factor
ggplot(data, aes(grp,
fill=factor(outcome,labels=c("double negative","positive for a")))) +
geom_bar() +xlab("group") +ylab("number of subjects") +
labs(fill="Serologic response")
http://stackoverflow.com/questions/7323191/how-do-i-manually-change-the-key-labels-in-a-legend-in-ggplot2
R factors, levels, labels
# sample data originally numbers
# 1 = male, 2 = female
people.n <- c(1, 1, 2, 2, 2)
people.c <- c("male", "male", "female", "female", "female")
# create a factor
people.n.f <- factor(people.n, levels = c(1, 2), labels = c("male", "female"))
people.c.f <- factor(people.c, levels = c("male", "female"), labels = c("male", "female"))
# if you want to change the order of the levels,
people.n.f <- factor(people.n, levels = c(2, 1), labels = c("female", "male"))
# levels follows alphabetical order if not specified
# compare this with people.c.f
factor(people.c)
# without specifying levels and labels
# the original data will be used
factor(people.n)
# once used to create a factor, levels and labels are merged.
# there is no labels attribute in factor.
# correct
levels(people.c.f)
levels(people.c.f) <- c("man", "woman")
people.c.f
#wrong
labels(people.c.f)
labels(people.c.f) <- c("man", "woman")
Conclusion
By default, factor() order the levels alphabetical order or numerical order.
If you want to change the sorting order of levels manually, you need to set it when the factor is created.
It can not be changed later, even though the levels can be changed.
When creating a factor, the data and the order should be matched as the example.
factor(people.c, levels = c("male", "female"), labels = c("male", "female"))
^ ---- same data ---------^ ^-------------same order -----^
Nice guide for R graphics
http://blog.revolutionanalytics.com/2009/01/10-tips-for-making-your-r-graphics-look-their-best.html
Treemap in R
1. Portpolio package
Pros: Quick, simple, bigger output
Cons: One color scheme
2. Treemap package
Pros: More customization, several algorithms
Cons: smaller output size
http://cran.r-project.org/web/packages/treemap/treemap.pdf
Remarks:
Index must be factor.
Different types
comp: (size – color) / color
dens: color / size
Customize the color, order and label of the legend in ggplot2
Use scale_colour_manual()
scale_colour_manual(values = c("red", "blue"), breaks = c("S2", "S1"), labels = c("sample1", "sample2"))
http://had.co.nz/ggplot2/scale_manual.html