Split-Apply-Combine (Divide and conquer) in a remote tibble or dbplyr class

Hi, I have this question

Does there exist any way to apply a "Split-Apply-Combine" (divide and conquer algorithm) method to a dbplyr object class? I found the Replyr package but it method wasn't work.

Currently I am connected with Rpostgres library to a Postgresql data base. So, in this way I am querying my data hosted in a postgres database, now just I need to apply the "Split-Apply-Combine" method (Divide and Conquer Algorithm) to get some results.

I am working in this way

fun_connect<-function(){dbConnect(RPostgres::Postgres(),dbname = 'database', 
                                  host = 'localhost', # i.e. 'ec2-54-83-201-96.compute-1.amazonaws.com'
                                  port = 5432, # or any other port specified by your DBA
                                  user = 'postgres',
                                  password = 'secretPassword',
                                  options="-c search_path=schema")}

#Activo funciĆ³n
conn <- fun_connect()

#Conecto y llamo a BBDD "censo"
dbtable<-tbl(conn, "table")

Replyr approach: dont'work

dbtable %>% replyr::replyr_split('columSplit') %>% 
  lapply(funToApply) %>% 

I am searching something like the local "Split-Apply-Combine" approach (this is just a prototype example):



split.data.frame(dfClassic, dfClassic$b)%>%

In the dbplyr github issues, the people recommend me use do() function, I tried the next:

dbtable%>% group_by("anycolumn") %>% do(mean(as.numeric("columnToApplyAFunction"))) %>% collect()

but this return me:

|======================================================================================= |100% ~0 s remaining
Error: No more ticks
Run `rlang::last_error()` to see where the error occurred.
In addition: There were 50 or more warnings (use warnings() to see the first 50)

Thanks for you help.

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.