在 R 的数据框中计算不同列中的独立值

Question

I have a data frame with several different columns and I want to get the frequency of different variables, check how the frequency of data changes depending on a parameter or two and compare the changes from the known or available ids to the missing ids where I have NA .

The objects Id is always known but there are cases where rq_ind is missing and those are interesting.

Basically I need to get the Nieseln area class of available object / Nielsen area of all ( missing and not missing objects ( which is the inq_onr_id==NA but their object_id is available)

    rq_id , rq_object_id , inq_onr_id,  inq_id,  Nielsen class, age_class,   revnue-class , employee_class                              
    157467  19750137    19750137    NA  3   3   4   2
    157467  19750137    19750137    NA  3   3   4   2
    423008  19750137    NA          NA  3   3   4   2   
    423008  19750137    NA          NA  3   3   4   2   
    157467  19750137    NA          NA  3   2   4   2   

    B1_fourth3month19short<-data.frame(rq_id,
                                     rq_object_id,
                                      inq_onr_id,inq_id,
                                      nielsen_area,Employeeclass)

All info are factors in principle.

What I want to take out is to find out how the fre(rq_object_id) vs Nieslen area changes in the case where the onr-id is missing vs where the onr _id is available.

Answer 1

What you need to use is table :

out <- table(df[,c(2,3,5)],useNA = "ifany")

where df is your initial data frame.

Output:

> out
, , Nielsen_class = 3

            inq_onr_id
rq_object_id 19750137 <NA>
    19750137        2    3

To get the percent of each row, do the following:

out.percent <- prop.table(table(df[,c(2,3,5)],useNA = "ifany"))*100

Output in percent:

> out.percent
, , Nielsen_class = 3

            inq_onr_id
rq_object_id 19750137 <NA>
    19750137       40   60