R- sqldf error raw vs double

Question

I have a vector lims, with the limits of a score:

 [1]  0.000000  7.025894  9.871630 12.411131 15.155998 18.099176 21.431354 25.391163 30.616550 40.356630

I create a table to classify other clients with:

lims[1]<- -0.00001 
a<-data.frame(lims[2:10])
colnames(a)<-'maxSc'
a<-rbind(a, 100)
lims<-data.frame(lims)
lims$maxSc<-a
colnames(lims)<-c('minSc', 'maxSc')


> class(lims)
[1] "data.frame"

So my result is:

> lims
       minSc      maxSc
1  -0.000010   7.025894
2   7.025894   9.871630
3   9.871630  12.411131
4  12.411131  15.155998
5  15.155998  18.099176
6  18.099176  21.431354
7  21.431354  25.391163
8  25.391163  30.616550
9  30.616550  40.356630
10 40.356630 100.000000

I want to classify another table (train) according to those limits and the field sc:

>class(train$sc)
[1] "numeric"
> class(train$target)
[1] "integer"

But when I run the following I get an error:

library(sqldf)
sqldf("Select b.minSc, b.maxSc, count(*) as total, sum(target) as compra
     from  train a left join lims b 
    on a.sc<=b.maxSc and a.sc>b.minSc
    group by b.minSc, b.maxSc")

Error in sqliteSendQuery(conn, statement, bind.data) : RAW() can only be applied to a 'raw', not a 'double'

I don't understand what I'm doing wrong.

Answer 1

I believe the error lies in your lims object.

lims <- c(0.000000,  7.025894,  9.871630, 12.411131, 
          15.155998, 18.099176, 21.431354, 25.391163, 
          30.616550, 40.356630)

lims[1]<- -0.00001 
a<-data.frame(lims[2:10])
colnames(a)<-'maxSc'
a<-rbind(a, 100) 
lims<-data.frame(lims)
lims$maxSc<-a
colnames(lims)<-c('minSc', 'maxSc')
sapply(lims, class)

#     minSc        maxSc 
# "numeric" "data.frame" 

Notice that lims$maxSc is of type data.frame. Then the following query doesn't work and results in the error you posted.

library(sqldf)
sqldf("select * from lims")

However, if instead lims$maxSc is set to a[,1] then there is no error.

lims$maxSc<-a[,1]
sapply(lims,class)
#     minSc     maxSc 
# "numeric" "numeric" 
sqldf("select * from lims")

The columns of your data.frame cannot be of class data.frame for sqldf to work.