In the case of multivariate analysis, we consider more than two variables for the study. The general approach is to perform single variable analysis and then, double variable analysis, and, finally, consider the significant one for multivariate analysis.
Let's first perform cross-tabulation with three variables. It is very similar to the two-variable cross-tabulation analysis but, instead of just two variables, we pass three variables here. Using the ftable function, we can see the tabular data:
tab<-xtabs(~Survived+Sex+SibSp, data=tdata) ftable(tab)
The output of the preceding command is as follows:
In order to get the details in percentages, we can use the following code. As there are three variables, we need to specify the column on which the percentage has to be computed. In the following code, we specify prop.table(tab, 3), which means that the percentages will be computed based on the third column, SibSp:
result <- replace(tab, , sprintf...
