Hello I have a simple question but cannot find easily the cl Hamilton Open Source #hamilton-help

Hello, I have a simple question but cannot find ea...

Paul D

10/14/2025, 2:32 PM

Hello, I have a simple question but cannot find easily the clean way to do it in the doc. How to rename a node to not have the function node. In particular, I'd like to have a space in the node name. Code would be like

Copy code

@rename("My Feature")
def my_feature(...) -> ...:
    return this_feature

And then node is named My Feature in the DAG, and I can further access it with

source("My Feature")

if needed.

Seth Stokes

10/14/2025, 4:12 PM

Check out this example. https://github.com/apache/hamilton/blob/343291bb99b56ab9093783f99940b9750d7788b9/examples/pandas/split-apply-combine/my_functions.py#L114

Copy code

# We use @inject decorator here because we have spaces in the names of columns.
# If column names are valid python variable names we wouldn't need this.
@inject(income=source("Income"), tax_rate=source("Tax Rate"), tax_credit=source("Tax Credit"))
# Step 4: 'Tax Formula' is calculated from 'Income', 'Tax Rate' and 'Tax Credit' series
def tax_formula(income: Series, tax_rate: Series, tax_credit: Series) -> Series:

👍 1

Stefan Krawczyk

10/14/2025, 4:20 PM

@Paul D you'd have to do it via

extract_fields

I think --> so return a dictionary and then pull that key out... I don't see a reason to not add

@rename

-- mind opening an issue / considering contributing it? The only challenge with this is that you're forced to use

@source

downstream because it ceases to be a valid python variable name.

Stefan Krawczyk

10/14/2025, 4:21 PM

If this issue is retaining spaces, I would consider a transformation on the final output of Hamilton?

Thierry Jean

10/14/2025, 5:55 PM

@Stefan Krawczyk I don't have specific example in mind, but I believe having nodes be valid identifiers is a strong assumptions made in several places in the code. For example, when using

@pipe_output

we namespace things

{entrypoint_node}.with_{step_node}

. Similar things happen in

@subdag

It creates a lot of edge cases if you can do

Copy code

@rename("; DROP TABLE bar")  # SQL injection
def foo(...) -> str: ...

@rename("<b><i>bingo<</b>")  # valid or invalid HTML
def baz(...) -> int: ...

also: • break graphviz notation for visualization • Node names are used as keys in dictionaries,

HamiltonGraph

• node names are passed to SQL engines in the caching backends and the Hamilton UI (would need further sanitization) • could even break exceptions and logging message if it includes

Thierry Jean

10/14/2025, 6:02 PM

A practical solution would be to have lifecycle like: 1. user node is @renamed

"My Node!"

2. pre-graph construction hook sanitizes nodes and maps

My Node!

my_node

3. error if we have name collisions (e.g., the user code has both

My Node!

and

My Node

and

my_node

already) 4. User can call

.execute()

using

"My Node!"

which internally maps to

my_node

.execute()

returns a dictionary

{"My Node!": ...}

## Edge cases / questions • what

node_name

do lifecycle hooks see? (e.g., caching, Hamilton UI, MLFlow plugin) • we need strong assumptions and sanitation for anything internal

Stefan Krawczyk

10/15/2025, 12:00 AM

rename could be restrictive

Stefan Krawczyk

10/15/2025, 12:00 AM

but yeah internally it is a string / string key -- and maybe graphviz would break.

Paul D

10/16/2025, 12:08 PM

I'd be happy to help if possible. I can open an issue where this can be further discussed

👍 1

Elijah Ben Izzy

10/19/2025, 4:58 PM

@rename

should be easy

Elijah Ben Izzy

10/19/2025, 4:58 PM

Problem is spaces in node name. I'd recommend keeping them as identifiers cause it makes things easier. Then add

@tag(name="My Exciting Name!")

and do the accounting yourself.

Open in Slack

Previous Next