Add support for int64 results

[edgararuiz-zz]

> tbl(con, in_schema("datawarehouse", "flight")) %>%
+     group_by(month) %>%
+     tally() %>%
+     collect() 
# A tibble: 12 x 2
   month               n
   <dbl> <S3: integer64>
 1    10          556205
 2     4          598126
 3    12          544958
 4     5          606293
 5     2          569236
 6     7          627931

[quartin]

Hi, I'm having the following problem (using dbplot_0.2.1.9000), which seems related to this issue:

> calls_ntiles_filter %>% 
+   head() %>% 
+   collect()
# A tibble: 6 x 2
  total_duration waiting_time
           <int>        <int>
1            161           72
2            141           72
3             93           72
4            634           72
5             71           72
6            410           72

> calls_ntiles_filter %>% 
+   dbplot_histogram(x = waiting_time)
Don't know how to automatically pick scale for object of type integer64. Defaulting to continuous.
Warning message:
Removed 29 rows containing missing values (position_stack). 

Any thoughts on why this is happening?

[quartin]

After looking at the PR that fixed this and later commits, seems this has been reverted?
edgararuiz@62be8b9#diff-e6b4f46f89e561a9db5fb9d2d83e8f26L47

[edgararuiz-zz]

Hi, thanks for reporting this. The reason why I reverted it is that I'm deferring the solution to ggplot2: tidyverse/ggplot2#2377

I may need to revert and add the work around until the fix is applied in ggplot2

[henningsway]

What is the suggested workaround for now, when the dataset is too big to be pulled into memory (and therefore cannot be coereced to numeric in R?)?

Temporarily transform the type of the column in the database via some sql command maybe?

[edgararuiz-zz]

Hi, you can use the corresponding db_ function in dbplot, which returns a regular data.frame, coerce that, and then pipe into a plot. Which plot are you trying to build?

[edgararuiz-zz]

@quartin / @henningsway - I reopened this issue because I just noticed that I may have inadvertently remove the temp fix I had to coerce automatically coerce the results.

[henningsway]

Hi, thanks for reporting this. The reason why I reverted it is that I'm deferring the solution to ggplot2: tidyverse/ggplot2#2377

I may need to revert and add the work around until the fix is applied in ggplot2

I think it would be really nice to be able to use the dbplot-plotting functions for int64-data, otherwise its less useful, if your database uses int64 types. I see, that handbuilding (computation, collect, coerce, plot) is an option, but it often takes four lines of code instead of one.

(And I suppose the dbplot-commands are more for the quick-analysis and general concept of "doing things in the database" anyways?)