crplyr 0.4.0

The first argument to autoplot and plotCategorical functions is renamed from x to object, for consistency with ggplot2’s autoplot generic.

crplyr 0.3.9

Converting CrunchCube results to the tibble like object “tbl_crunch_cube” is now done via as_cr_tibble() and as_tibble() always returns an actual tibble.
Updates for upcoming dplyr 1.0 release

New autoplot() methods make it easy to plot crunch variables and cubes as well as cube calculations (proportions from prop.table() margins from margin.table(), etc.). See vignette("plotting", package="crplyr") for discussion and examples.
as_tibble() and summarize() return full underlying cube representations, where multiple-response items are represented as an array of multiple dichotomous choices (selected, not-selected, missing) for each item.
select(), group_by(), and collect() now correctly support taking hidden variables (#6, #13)
collect() uses crunch::as.data.frame() export, which should be faster especially for larger data pulls.
Internal methods summarize_(), select_(), and group_by_(), deprecated in dplyr, now error.
Improved documentation and error messages (#5, #9).

Added collect() method, which pulls the requested columns of data from the server.
summarize() and as_tibble.CrunchCube() methods now better handle array and multiple-response data and include metadata on which dimension values should be interpreted as missing.
unweighted_n() aggregation method for summarize(), which returns the unweighted counts even when the dataset has a weight applied.

Initial implementation of dplyr interface for Crunch, including

select(), filter(), group_by(), and summarize() methods for crunch::CrunchDataset() objects
as_tibble() method for crunch::CrunchCube() objects