Clase R @ LIM 2014

clase_R_2014.R

a <- c(4, 6, 3, 7, 3, 6, 3, 0)
mean(a)
sd(a)
median(a)
var(a)
a[1]
a[1:4]
a[c(1, 3, 8)]

# Comentario. Esto no se ejecuta.
a[1] <- 5
a
a[2] <- NA
a
mean(a)
mean(a, na.rm = TRUE)

class(a)
b <- c("a", "c", "d")
class(b)

class(iris)
length(a)
length(iris)
dim(iris)
nrow(iris)
ncol(iris)
head(iris)
?head
head(iris, n = 10)
summary(iris)
mean(iris$Sepal.Length)
class(iris$Species)

# Factores
factorEx <- c(rep(1, 10), rep(2, 10), rep(3, 10))
factorEx2 <- factor(factorEx, labels = c("grupo1", 
                                         "grupo2",
                                         "grupo3"))
head(iris)
iris[1, ]
iris[1, c(1, 3)]
iris[1, -5]
iris[1, -c(1, 5)]
testIris <- iris[iris$Species == "versicolor", ]
index1 <- iris$Species == "versicolor"
testIris <- iris[index1, ]
testIris <- iris[!index1, ]
TRUE | FALSE
TRUE & FALSE
TRUE & !FALSE
which(index1)

# Bucles y estructuras de control
for (i in a) {
  print(i)
}

for (i in 1:length(a)) {
  print(a[i])
}

for (i in 1:length(a)) {
  if (a[i] < 7) {
    print(a[i])
  }
}

# Funciones
f1 <- function(x, y = 5) {
  ret <- list(orig = x, 
              mod = x - y,
              m = mean(x, na.rm = TRUE))
  return(ret)
}
b <- f1(1:10)
class(b)
b$orig
b$mod
b2 <- f1(1:10, 2)

iris2 <- iris[, -5]
summary(iris2)
colMeans(iris2)
rowMeans(iris2)
colSums(iris2)

apply(iris2, 2, sd)
sapply(iris2, sd)

iris2[20, 3] <- NA
apply(iris2, 2, sd, na.rm = TRUE)
apply(iris2, 2, function(x) sd(x, na.rm = TRUE))

for (i in 1:ncol(iris2)) {
  print(sd(iris2[, i], na.rm = TRUE))
}

aggregate(Sepal.Length ~ Species, data = iris, mean)

a <- read.table(url("http://core2.gsfc.nasa.gov/research/purucker/mgs_gridded.data"))
names(a)
names(a)[1] <- "Variable1" 
names(a)
a <- read.csv(url("http://pims.grc.nasa.gov/plots/user/mike/2013_05_sensor_gmt_excelt_medfreq_medrms.csv"))


install.packages("xlsx")
library(xlsx)
install.packages("foreign")
library(foreign)

# T tests, regresiones, etc
x <- rnorm(100)
y <- rnorm(100)
t1 <- t.test(x, y)
t.test(x, y, paired = TRUE, var.equal = TRUE)

x <- 1:100
y <- x + rnorm(100, sd = 20)
plot(x, y)
lm1 <- lm(y ~ x)
lm1
abline(lm1)
s1 <- summary(lm1)

for(i in 1:100) {
  x <- rnorm(100)
  y <- rnorm(100)
  lm1 <- lm(y ~ x)
  s1 <- summary(lm1)
  p1 <- s1$coefficients[2, 4]
  if (p1 < 0.05) {
    print(p1)
  }
}

x1 <- rnorm(100)
x2 <- rnorm(100)
y <- x1 + x2 + rnorm(100)
lm2 <- lm(y ~ x1 + x2)
summary(lm2)

# ANOVA y relacionados
lm3 <- lm(Sepal.Length ~ Species, data = iris)
anova(lm3)
TukeyHSD(aov(Sepal.Length ~ Species, data = iris))
pairwise.t.test(iris$Sepal.Length, iris$Species, p.adjust.method = "bonferroni")
shapiro.test(iris$Sepal.Length)
pairwise.wilcox.test(iris$Sepal.Length, iris$Species)
kruskal.test(Sepal.Length ~ Species, data = iris)

plot(iris$Petal.Length ~ iris$Species)
plot(density(iris$Petal.Length))
plot(iris)

with(iris, plot(Sepal.Length, Sepal.Width, col = as.numeric(Species)))


library(ggplot2)
theme_set(theme_bw())
ggplot(iris) + geom_point(aes(x = Sepal.Width, y = Sepal.Length, size = Petal.Length, color = Species))
ggplot(iris) + geom_point(aes(x = Sepal.Width, y = Sepal.Length)) + facet_wrap( ~ Species)
# Para multiplicar matrices: %*%