inject

Reference Documentation

class hamilton.function_modifiers.inject(**key_mapping: ParametrizedDependency)[source]

@inject allows you to replace parameters with values passed in. You can think of it as a @parameterize call that has only one parameterization, the result of which is the name of the function. See the following examples:

import pandas as pd
from function_modifiers import inject, source, value, group

@inject(nums=group(source('a'), value(10), source('b'), value(2)))
def a_plus_10_plus_b_plus_2(nums: List[int]) -> int:
    return sum(nums)

This would be equivalent to:

@parameterize(
    a_plus_10_plus_b_plus_2={
        'nums': group(source('a'), value(10), source('b'), value(2))
    })
def sum_numbers(nums: List[int]) -> int:
    return sum(nums)

Something to note – we currently do not support the case in which the same parameter is utilized multiple times as an injection. E.G. two lists, a list and a dict, two sources, etc…

This is considered undefined behavior, and should be avoided.

__init__(**key_mapping: ParametrizedDependency)[source]

Instantiates an @inject decorator with the given key_mapping.

Parameters:

key_mapping – A dictionary of string to dependency spec. This is the same as the input mapping in @parameterize.