Can't figure out how to add a summary function to count the rows in a group...
Try using the length() function.
With the following:
you are attempting to place the summary result in a column that is hidden. Try using columns = vars(num) instead.
Thanks for the workaround. As you note, it works, except that the length function only works if the your dataset has a disposable numeric variable to use with the columns parameter, and the count comes out with two trailing zeros.
As you work on a fix, I'd suggest that counting the rows in a group would be a common use case for a summary_rows function.
This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.
I got length() to work as a grand summary:
library(tidyverse)
library(gt)
exibble %>% select(group, char, date, num) %>%
gt(groupname_col = "group") %>%
summary_rows(
groups = NULL,
columns = vars(num),
fns = list(len = ~length(.)
))
| char | date | num | |
|---|---|---|---|
| grp_a | |||
| apricot | 2015-01-15 | 1.111e-01 | |
| banana | 2015-02-15 | 2.222e+00 | |
| coconut | 2015-03-15 | 3.333e+01 | |
| durian | 2015-04-15 | 4.444e+02 | |
| grp_b | |||
| NA | 2015-05-15 | 5.550e+03 | |
| fig | 2015-06-15 | NA | |
| grapefruit | NA | 7.770e+05 | |
| honeydew | 2015-08-15 | 8.880e+06 | |
| len | — | — | 8.00 |
Created on 2020-04-13 by the reprex package (v0.3.0)
But when groups = T, no summary is produced:
library(tidyverse)
library(gt)
exibble %>% select(group, char, date, num) %>%
gt(groupname_col = "group") %>%
summary_rows(
groups = T,
columns = vars(group),
fns = list(len = ~length(.)
))
| char | date | num |
|---|---|---|
| grp_a | ||
| apricot | 2015-01-15 | 1.111e-01 |
| banana | 2015-02-15 | 2.222e+00 |
| coconut | 2015-03-15 | 3.333e+01 |
| durian | 2015-04-15 | 4.444e+02 |
| grp_b | ||
| NA | 2015-05-15 | 5.550e+03 |
| fig | 2015-06-15 | NA |
| grapefruit | NA | 7.770e+05 |
| honeydew | 2015-08-15 | 8.880e+06 |
Created on 2020-04-13 by the reprex package (v0.3.0)
It looks like there is a bug where the summary rows don't get built if there is no stub. I'll file an issue shortly to ensure this gets fixed. As a temporary workaround, this should work:
library(tidyverse)
library(gt)
exibble %>%
select(group, char, date, num) %>%
mutate(stub = rep(" ", nrow(.))) %>%
gt(groupname_col = "group", rowname_col = "stub") %>%
summary_rows(
groups = TRUE,
columns = vars(num),
fns = list(len = ~length(.)
)
)
That creates a column with space characters before using gt(), then that stub column is specified in the gt() call.
No, I tried that:
library(tidyverse)
library(gt)
exibble %>% select(group, char, date, num) %>%
gt(groupname_col = "group") %>%
summary_rows(
groups = T,
columns = vars(num),
fns = list(len = ~length(.)
))
| char | date | num |
|---|---|---|
| grp_a | ||
| apricot | 2015-01-15 | 1.111e-01 |
| banana | 2015-02-15 | 2.222e+00 |
| coconut | 2015-03-15 | 3.333e+01 |
| durian | 2015-04-15 | 4.444e+02 |
| grp_b | ||
| NA | 2015-05-15 | 5.550e+03 |
| fig | 2015-06-15 | NA |
| grapefruit | NA | 7.770e+05 |
| honeydew | 2015-08-15 | 8.880e+06 |
Created on 2020-04-13 by the reprex package (v0.3.0)