What does the error Error in `stop_subscript()`mean when applying knn function from the neighbr package?

[JuliaG]

Hi, I keep getting the following error when I run the knn function from the neighbr package:

My intention is to use the knn function for unsupervised learning given the example in the help section of the package, so without a target. Here is my code:

fit<-knn(train_set=train_set, test_set = test_set,
k=5,
comparison_measure = "euclidean",
return_ranked_neighbors = 2,
id="ID")

My data is a mix of 6 binary variables and 2 continues standardized ones. The train set also contains an ID column.
The train set contains 5000 rows and the test set 990.
Is the problem in my data or with the way I use the function?

Thanks for help!

[technocrat]

See the FAQ: How to do a minimal reproducible example reprex for beginners for how to attract more targeted answers.

If ID is character, try with a copy of the data.frame that omits it.

[JuliaG]

Thank you for your reply.
My ID column is an integer. I tried using the function with the id, but I get an error saying that the ID must be specified:
Error in knn(train_set = train_set, test_set = test_set, k = 5, comparison_measure = "euclidean", :
if return_ranked_neighbors > 0, id column must be specified.
Is there any other solution?

[technocrat]

Without the reprex, there’s nothing I can help with.

[JuliaG]

Below is now the reprex. I hope this is right as the ID column has the values below with an L (I created the column using the rowid_to_column() function. However, with just the small sample the function works without an error. I still get the error when I use the full dataset. why?

df_train<-data.frame(stringsAsFactors = FALSE,
        check.names = FALSE,
                            ID = c(1L,2L,
                                   3L,4L,5L,6L,7L,8L,9L,10L,11L,12L,
                                   13L,14L,15L,16L,17L,18L,19L,20L),
                   BD_research = c(0,1,0,
                                   1,0,0,0,1,1,0,0,0,0,0,0,0,0,0,
                                   0,0),
                         BD_RD = c(0,0,0,
                                   0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,
                                   0,0),
                    BD_develop = c(0,1,0,
                                   1,0,0,1,1,1,1,1,0,1,1,1,0,0,1,
                                   1,0),
                   BD_discover = c(0,0,0,
                                   0,0,0,0,0,0,0,1,0,0,0,0,0,0,1,
                                   0,0),
             `Private Company` = c(1,1,0,
                                   0,1,1,1,0,0,1,0,0,0,0,0,1,1,1,
                                   1,1),
              `Public Company` = c(0,0,1,
                                   1,0,0,0,1,1,0,1,1,1,1,1,0,0,0,
                                   0,0),
              Number_Employees = c(-0.11129043029361,-0.133882618992182,
                                   -0.152212130577816,-0.0599251710827047,
                                   -0.0957316588313852,-0.0386117855180139,-0.0955185249757383,
                                   -0.142194839362411,-0.123225926209836,
                                   -0.143899910207586,-0.130472477301831,
                                   -0.0490553444447124,-0.148162587320525,
                                   -0.14432617791888,-0.148588855031818,-0.142407973218058,
                                   -0.110437894871022,-0.148375721176171,
                                   -0.0571544309592949,-0.0733526039884599),
                 Total_Revenue = c(-0.0980542624611477,-0.100277064165423,
                                   -0.111073529586189,-0.0958314607568724,
                                   -0.0770964178208376,0.019754227865444,-0.0993244348635908,
                                   -0.112026158888021,-0.11170861578741,
                                   -0.110438443384967,-0.112026158888021,
                                   -0.0755087023177838,-0.112026158888021,
                                   -0.112026158888021,-0.112026158888021,-0.109803357183746,
                                   -0.0888455125434357,-0.111073529586189,
                                   -0.0929735728513756,-0.0466122801622047)
           )

df_test<-data.frame(stringsAsFactors = FALSE,
        check.names = FALSE,
        BD_research = c(0, 1, 0, 0, 0, 0),
              BD_RD = c(0, 0, 0, 0, 0, 0),
         BD_develop = c(0, 1, 1, 0, 0, 0),
        BD_discover = c(0, 0, 0, 0, 0, 0),
  `Private Company` = c(1, 1, 0, 1, 1, 1),
   `Public Company` = c(0, 0, 1, 0, 0, 0),
              Number_Employees = c(-0.131751280435713,-0.13814529610512,
                                   0.417068397855075,-0.136013957548651,-0.153064666000403,
                                   -0.152851532144757),
                 Total_Revenue = c(-0.104087581372752,-0.108533184781303,
                                   0.479874180650434,-0.105040210674584,
                                   -0.0932911159519863,-0.112026158888021)
           )

library(neighbr)
#> Warning: package 'neighbr' was built under R version 4.1.3

f<-knn(train_set=df_train, test_set = df_test,
         k=3,
         comparison_measure = "euclidean",
         return_ranked_neighbors = 2,
         id="ID")

^{Created on 2022-12-17 with reprex v2.0.2}
Thanks again for your help!

[technocrat]

Don't worry about the L suffix—it just means typeof integer.

Assuming the same data layout in the full dataset, check for missing values, probably in column ID. You can get the row index for all variables with

complete.cases(full_data)

[quote="JuliaG, post:3, topic:155206"]
if return_ranked_neighbors > 0, id column must be specified.
[/quote]

There's no id column in df_test. Is there one in the test set for the full data?

[JuliaG]

My problem got fixed by adjusting the return_ranked_neighbors to only 1. For some reason my data did not work with a higher neighbors number.
Thanks again for your help.