Veg <- read.table(file = "Vegetation2.txt", header = TRUE)Verify that it is read in correctly
names(Veg)
str(Veg)
head(Veg)m <- mean(Veg$R)
m1<- mean(Veg$R[Veg$Transect == 1])
m2<- mean(Veg$R[Veg$Transect == 2])
m3<- mean(Veg$R[Veg$Transect == 3])
m4<- mean(Veg$R[Veg$Transect == 4])
m5<- mean(Veg$R[Veg$Transect == 5])
m6<- mean(Veg$R[Veg$Transect == 6])
m7<- mean(Veg$R[Veg$Transect == 7])
m8<- mean(Veg$R[Veg$Transect == 8])
c(m, m1, m2, m3, m4, m5, m6, m7, m8) The variable m contains the mean richness of all 8 transects, and m1 through m8 show the mean richness values per transect. Note that the mean command is applied to Veg$R, which is a vector of data. It is not a matrix; hence there is no need for a comma between the square brackets.
The R function tapply performs the same operation as the code above (for m1 through m8 ), but with a single line of code
tapply(Veg$R, Veg$Transect, mean)or
tapply(X = Veg$R, INDEX = Veg$Transect, FUN = mean)The tapply function splits the data of the first variable (R),based on the levels of the second variable (Transect). To each subgroup of data, it applies a function, in this case the mean, but we can also use the standard deviation (function sd()), variance (function var()), length (function length()),and so on. The following lines of code calculate some of these functions for the vegetation data.
Me <- tapply(Veg$R, Veg$Transect, mean)
Sd <- tapply(Veg$R, Veg$Transect, sd)
Le <- tapply(Veg$R, Veg$Transect, length)
cbind(Me, Sd, Le) Each row in the output gives the mean richness, standard deviation, and number of observations per transect.
To calculate the mean, minimum, maximum, standard deviation, and length of the full series, we still need to use mean(Veg$R), min(Veg$R), max(Veg$R), sd(Veg$R), and length(Veg$R). This is laborious if we wish to calculate the mean of a large number of variables such as all the numerical variables of the vegetation data. We specifically say 'numerical' as one cannot calculate the mean of a factor. There are 20 numerical variables in the vegetation dataset, columns 5–25 of the data frame Veg. However, we do not need to type in the mean command 20 times. R provides other functions similar to the tapply to address this situation: lapply() and sapply().
sapply(Veg[, 5:9], FUN= mean)It is important to realise that tapply calculates the mean (or any other function) for subsets of observations of a variable, whereas lapply and sapply calculate the mean (or any other function) of one or more variables, using all observations.
The word FUN stands for function, and must be written in capitals. Instead of the mean, you can use any other function as an argument for FUN, and you can write your own functions.
So, what is the difference between sapply() and lapply()? The major differences lie in the presentation of output, as can be seen in the following example.
lapply(Veg[ , 5:9], FUN = mean)The output of lapply is presented as a list, whereas sapply gives it as a vector. The choice depends on the format in which you would like the output. The variable that contains the data in lapply and sapply needs to be a data frame. This will not work:
sapply(cbind(Veg$R, Veg$ROCK, Veg$LITTER, Veg$ML, Veg$BARESOIL), FUN = mean)It will produce one long vector of data, because the output of the cbind command is not a data frame. It can easily be changed to a data frame:
sapply(data.frame(cbind(Veg$R, Veg$ROCK, Veg$LITTER, Veg$ML, Veg$BARESOIL)), FUN = mean)Note that we have lost the variable labels. To avoid this, make a proper data frame before running the sapply function. Alternatively, use the colnames function after combining the data with the cbind() function.
Another function that gives basic information on variables is the summary() command. The argument can be a variable, the output from a cbind() command, or a data frame. It is run by the following commands.
Z <-cbind(Veg$R, Veg$ROCK, Veg$LITTER)
colnames(Z) <- c("R", "ROCK", "LITTER")
summary(Z)The summary command gives the minimum, first quartile, median, mean, third quartile, and maximum value of the variable. An alternative R code gives the same result:
summary(Veg[ , c("R","ROCK","LITTER")])or summary(Veg[ , c(5, 6, 7)])
The table() function can be used to learn how many animals per farm were sampled, as well as the number of observations per sex and year. The following code imports the data, and tells you the attributes of the data.
Deer <- read.table(file = "Deer.txt", header = TRUE)
names(Deer)
str(Deer) Look at the first few lines of the data
head(Deer)Farm has been coded as AL, AU, and so on, and is automatically imported as a factor. The other variables are all vectors of numerical or integer values. The number of observations per farm is obtained by
table(Deer$Farm)
table(Deer$Sex, Deer$Year)The file temperature.xls contains temperature observations made at 31 locations along the Dutch coastline. The data were collected and provided by the Dutch institute RIKZ (under the monitoring program MWTL; Monitoring Waterstaatkundige Toestand des Lands). Sampling began in 1990, and the final measurements in the spreadsheet were taken in December 2005, a period of 16 years. Sampling frequency was 0–4 times per month, depending on the season.