age nr.employed          job  marital           education housing default emp.var.rate cons.price.idx cons.conf.idx
1   30      5099.1  blue-collar  married            basic.9y     yes      no         -1.8         92.893         -46.2
2   39      5191.0     services   single         high.school      no      no          1.1         93.994         -36.4
3   25      5228.1     services  married         high.school     yes      no          1.4         94.465         -41.8
4   38      5228.1     services  married            basic.9y unknown      no          1.4         94.465         -41.8
5   47      5195.8       admin.  married   university.degree     yes      no         -0.1         93.200         -42.0
6   32      4963.6     services   single   university.degree      no      no         -1.1         94.199         -37.5
7   32      4963.6       admin.   single   university.degree     yes      no         -1.1         94.199         -37.5
8   41      5195.8 entrepreneur  married   university.degree     yes unknown         -0.1         93.200         -42.0
9   31      5195.8     services divorced professional.course      no      no         -0.1         93.200         -42.0
10  35      5191.0  blue-collar  married            basic.9y      no unknown          1.1         93.994         -36.4

 

In a number of variables of the data frame survey the string "unknown" is used to indicate a missing value. Replace all these instances with proper missing values. Keep in mind that you’re changing values of factors, so you also have to adjust the levels! Store your result in a data frame called survey_missing