You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This isn't necessarily limited to CSVs, but we should have this philosophy in our developing vignette
The general approach we've done with CSV things in the past is:
(1) support the readr signature to the best extent we can, translating to the arrow parameter names internally;
(2) allow someone to pass the arrow options (CsvReadOptions etc.) directly, in case they want to do things at a lower level
(3) where necessary add extra args to the readr signature for features that don't exist in R but do in arrow (e.g. schema)
This is our general philosophy: present things to the R user in a way that is least surprising to them (most follows R conventions) and also provide access to all of Arrow's features (sometimes that's extra args, sometimes it's the arrow_ prefixed functions in dplyr, sometimes it's just the lower-level R6 objects and methods that more closely follow the C++ interface)
Nicola Crane / @thisisnic:
I am currently working on a document around writing compute bindings, and this feels like it'd fit alongside that. I'm wondering if, instead of 1 long vignette, we need a series of shorter articles for the pkgdown site, like in this example? https://pkgdown.r-lib.org/reference/build_articles.html
This isn't necessarily limited to CSVs, but we should have this philosophy in our developing vignette
The general approach we've done with CSV things in the past is:
(1) support the readr signature to the best extent we can, translating to the arrow parameter names internally;
(2) allow someone to pass the arrow options (CsvReadOptions etc.) directly, in case they want to do things at a lower level
(3) where necessary add extra args to the readr signature for features that don't exist in R but do in arrow (e.g. schema)
This is our general philosophy: present things to the R user in a way that is least surprising to them (most follows R conventions) and also provide access to all of Arrow's features (sometimes that's extra args, sometimes it's the arrow_ prefixed functions in dplyr, sometimes it's just the lower-level R6 objects and methods that more closely follow the C++ interface)
Reporter: Jonathan Keane / @jonkeane
Assignee: Nicola Crane / @thisisnic
PRs and other links:
Note: This issue was originally created as ARROW-14441. Please see the migration documentation for further details.
The text was updated successfully, but these errors were encountered: