PipeableCallCheckWarning: Failed to detect AST node calling `xxx`, assuming a normal call

[pwwang]

If you see this warning and the errors caused by it, please ensure you are running your code in the environment where the source code is available at runtime.

Typical environment where the source code is not available at runtime:

• Raw Python REPL (type python from command line and run your code there)
• exec()

If you want to use a Python REPL to run your code, ipython, bpython and jupyter notebook are recommended.

If you can't avoid using exec(), use exec_code() from varname.helpers instead.

What if I want to use Raw Python REPL or exec() anyway? The idea is to tell datar if you are using "piping" or "normal" calling for the function. You can choose either way:

• Set ast_fallback for a function
• Pass __ast_fallback to a function call

from datar.all import f, select
from datar.data import iris

iris >> select(f.Species)  # warning

select.ast_fallback = "piping"
iris >> select(f.Species)  # ok
# but if you call it "normally"
select(iris, f.Species)  # Unevaluated `VerbCall object`

# or you can pass `__ast_fallback`
iris >> select(f.Species, __ast_fallback="piping")  # ok
select(iris, f.Species, __ast_fallback="normal")  # ok

Please also refer to:

• https://pwwang.github.io/pipda/piping/
• [BUG] warnings when using pipe #202
• [QST] Warning: PipeableCallCheckWarning: Failed to detect AST node calling mutate, assuming a piping call. #196
• warning #198
• Failed to detect AST nodes in Windows ipython for verb calls #170
• piping to verbs is not passing the dataframe argument #149
• dataframe to tibble #55
• pipe operator doesn't work in plain python prompt #45