option to clean/normalize data.frames?

[krlmlr]

Fixes tidyverse/dplyr#1587.

@r2evans: Would you like to contribute to this package?

[r2evans]

What's the purpose of the package? I haven't gone through the roxy-docs for the package yet, what's the purpose for tibble? (I don't mind contributing to the package, but the description is a little brief.)

[krlmlr]

Eventually, dplyr will import from tibble. This package extracts the functionality around tbl_df from dplyr to make it available for other packages such as tidyr and readr, see also https://github.com/krlmlr/tibble/releases/tag/v0.1 and tidyverse/dplyr#1488.

[r2evans]

Perfect, I understand. Hadley was going to mull over the "if" and "where" of the suggested patch (from hadley/dplyr#1587). Did you have any thoughts on it? I'll be happy to create the PR for it.

[krlmlr]

How about fix_names() or even repair_names()? I think that for any data frame where tbl_df() raises an error, repair_names() %>% tbl_df() should just work -- this is what also should be tested in the unit tests. Data frames for which tbl_df() already works should not be touched.

We can work on this together and discuss the results with Hadley.

[krlmlr]

@r2evans: Are you still interested in contributing? This could be useful for as_data_frame.matrix():

> dplyr::as_data_frame(diag(3))
Source: local data frame [3 x 3]

     NA    NA    NA
  (dbl) (dbl) (dbl)
1     1     0     0
2     0     1     0
3     0     0     1

[r2evans]

Yes, thanks for the poke. This is a good test case, too.