I would like to change a class defined by two values in brackets by the mean value of the corresponding classes. Here is the situation: I am using "cut" to split a variable into 10 classes.
data$classe_Ta<-cut(data$Ta,10,include.lowest = TRUE)
table(data$classe_Ta)
The result gives:
> [-12.4,-7.81] (-7.81,-3.25] (-3.25,1.31] (1.31,5.87] (5.87,10.4] (10.4,15] (15,19.6] (19.6,24.1] (24.1,28.7] (28.7,33.3]
59 490 2783 6028 7561 7051 5090 1400 211 15
I would like to replace [-12.4,-7.81], (-7.81,-3.25] ...in my data frame by the mean of each class; which means by -10.105, -5.53....which needs to be calculated for each class. Any suggestions in how to do that would be much appreciated. Thanks in advance!
We could use str_extract to extract all the numeric elements, convert the string to numeric, get the mean.
v1 <- sapply(str_extract_all(levels(data$classe_Ta),
"-?[0-9]+(\\.[0-9]+)?"), function(x) mean(as.numeric(x)))
data$Mean <- v1[as.numeric(data$classe_Ta)]
set.seed(24)
data <- data.frame(Ta = sample(-50:100, 20, replace=TRUE))
data$classe_Ta<-cut(data$Ta,10,include.lowest = TRUE)