Michael Koohafkan

This issue also causes inconsistent handling of NA values, i.e. the behavior of `read_excel(na = "")` is different from `read_excel(na = "NA")` or even `read_excel(na = " ")` ``` #...

Exactly. But in the first example we read in empty cells where the `NA` match is `""` (empty) so I would expect to get NA values back in that case...

Agreed, but I wasn't sure what the preferred convention was or if there was a specific reason for the result of the first test case. More generally, my initial expectation...

right now I'm reading excel files from a URL using ```r tf = tempfile(fileext = ".xlsx") curl::curl_download(url, tf) readxl::read_excel(tf, ...) ``` But I could also see a more generic approach...

This feature could potentially resolve a separate but related issue when hoisting a zero-row tibble: ```r library(dplyr) library(tidyr) # no hoist column is returned at all mtcars %>% group_nest(cyl) %>%...

The behavior is different because you are doing different things. In the first case, `map_dbl(1:4, ~ one())` is roughly the same as ```r for (i in 1:4) one() ``` whereas...

The purpose of `map()` is to iterate a function over a list or vector of argument values. This is similar to the `*apply()` family of functions, which also fail with...

From the documentation of `purrr::map()`: > The map functions transform their input by applying a function to each element of a list or atomic vector and returning an object of...

I think the best solution is to rename `igraph::as_data_frame` to **`igraph::to_data_frame`**. Makes it more consistent with `igraph::from_data_frame` and also is a clue that the behavior is fundamentally different from `dplyr::as_data_frame`...

I have run into this as well. Current workaround is to do `do.call(union, list.of.graphs)`. I don't know how one could ever get this function to work as intended with a...