Eugene Ha

This affects indexing semantics. Currently, `((f) %>>>% g)[[1]]` is always `f`, but `(f %>>>% g)[[1]]` is the first function in the flattened composite, which is not necessarily `f` (because `f`...

Printing in Jupyter notebook

4

Executing a cell like ```R abs %>>>% log ``` ignores the class and prints ``` structure(function (x) `__2__`(`__1__`(x)), class = c("CompositeFunction", "function" )) ``` whereas ```R print(abs %>>>% log) ```...

Methods of type `Posure -> Posure` for manipulating the underlying composite function, safely. (Note: The posure expression itself can be manipulated by privately overloading `%>>>%` as a concatenation action on...

Functional for creating posures

1

Currently, `posure()` acts on a function declaration whose expression qua body must be a `%>>>%` chain. The problem is that a function declaration is not a type. A safer and...

Accordingly, `%>>>%` is the _chain_ operator; use this terminology in the docs and README.

Bring back `as_fn()`, in modified form: ```R as_fn(x, y, ... ~ body) # fn(x, y, ... ~ body) as_fn(~ foo(.)) # fn(. ~ foo(.)) as_fn(..f = foo) # foo as_fn(..f...

- Demonstrate use and purpose of new features of 0.1.4 (`posure()`, `let()`, `run()`) - Clearly articulate the complementary roles of `%>>>%` and `%>%` - Real-world examples of composition

`do()` is a more fitting name than `run()`. (Note that this would conflict with `dplyr::do()`.)