A tip I wanted to share: You can name a node using the dunder method

__name__

instead of typing out a

str

.

node(
func=explode_missing_values,
...
name=explode_missing_values.__name__,
)

# is equivalent to

node(
func=explode_missing_values,
...
name="explode_missing_values",
)

If you're setting the node name to be the function name, this makes it so renaming the function and the node is just one step.

out of interest, would you like this if it was default behaviour? i.e. if not provided we do this

Just tried commenting out where I named the nodes and am wondering if this is default behavior already? I think it would be a good default behavior, if it's not already.

I think the default is

name=None

, I'm 100 % supportive of making

func.__name__

the default if not given

Would this count as a breaking change?

yeah, better to proceed with care

my point is we can’t squeeze it into 0.19.x right?

From what I can tell there is no difference in Kedro Viz between

name=None

and

name=func.__name__

but the logger shows a minor difference: None:

Running node: explode_missing_values([input]) -> [output]

___name___:

Running node: explode_missing_values: explode_missing_values([input]) -> [output]

Yeah I think the point is Viz is inferring this rather than it really being true in the Python node

my point is we can’t squeeze it into 0.19.x right?

I think it's safe to say we're in "feature freeze" mode atm, the release will come in a few days and we're squashing the last blockers

shame - but we should put it in the next version

Thx @Chris Schopp Very handy indeed. And (for what my opinion is worth 😅 )I will second @Juan Luis in saying that too am “_100 % supportive of making

func.__name__

the default if not given_”

from the 0.19.0 release notes:

Anonymous nodes are given default names of the form

<function_name>([in1;in2;...]) -> [out1;out2;...]

, with the names of inputs and outputs separated by semicolons.