in_schema returns a list of 2 not a tbl

user_1024586 0
Viewed 234

I have a connection to our database:

con <- dbConnect(odbc::odbc(), "myHive")

I know the con is working because I can see all of our schemas in the upper right connections pane.

If I want to select a particular table as a tbl I can go:

mytbl <- tbl(con, in_schema("mydb", "mytable"))

I expected the reult here to be a df or tbl however it's a list of 2. I was trying to extract the tbl part but am getting confused by this list.

Here is the str:

> str(mytbl)
List of 2
 $ src:List of 2
  ..$ con  :Formal class 'Hive' [package ".GlobalEnv"] with 4 slots
  .. .. ..@ ptr     :<externalptr> 
  .. .. ..@ quote   : chr "`"
  .. .. ..@ info    :List of 13
  .. .. .. ..$ dbname               : chr "HIVE"
  .. .. .. ..$ dbms.name            : chr "Hive"
  .. .. .. ..$ db.version           : chr "1.2.2"
  .. .. .. ..$ username             : chr ""
  .. .. .. ..$ host                 : chr ""
  .. .. .. ..$ port                 : chr ""
  .. .. .. ..$ sourcename           : chr "gdHive"
  .. .. .. ..$ servername           : chr "Hive"
  .. .. .. ..$ drivername           : chr "Hortonworks Hive ODBC Driver"
  .. .. .. ..$ odbc.version         : chr "03.52"
  .. .. .. ..$ driver.version       : chr "2.6.1.1001"
  .. .. .. ..$ odbcdriver.version   : chr "03.80"
  .. .. .. ..$ supports.transactions: logi FALSE
  .. .. .. ..- attr(*, "class")= chr [1:3] "Hive" "driver_info" "list"
  .. .. ..@ encoding: chr ""
  ..$ disco: NULL
  ..- attr(*, "class")= chr [1:3] "src_dbi" "src_sql" "src"
 $ ops:List of 2
  ..$ x   : 'ident_q' chr "mydb.mytable"
  ..$ vars: chr [1:188] "zzz1.order_id" "zzz1.row_id" "zzz1.order_ts" "zzz1.order_date" ...
  ..- attr(*, "class")= chr [1:3] "op_base_remote" "op_base" "op"
 - attr(*, "class")= chr [1:4] "tbl_dbi" "tbl_sql" "tbl_lazy" "tbl"

What am I looking at here? How can I get the table from mydb.mytable as a tbl?

dbplyr dplyr r
1 Answers

If you are looking to load the table into R memory (rather than work with it remotely) use mytbl %>% collect().

Otherwise, I think what you are looking at is a standard remote table. This is not particular to in_schema. If you try class(mytbl) you should expect to see "tbl" as one of its classes.

By default R does not load data from a remote table into memory. To view the first few rows try: mytbl %>% head().

You can manipulate the table using all the standard dplyr commands. For example:

results = mytbl %>%
  rename(new_name = old_name) %>%
  mutate(new_col = 2*old_col) %>%
  group_by(new_col) %>%
  summarise(number = n()) %>%
  filter(number > 1000)

And then when you want the result in R: results = results %>% collect().

user_77429810
Related