@Marvin In prefect.yaml there is the pull section. What is the equivalent in python when using deploy()

In Prefect 3.x, the

pull

section in the

prefect.yaml

file is used to define how your deployment is pulled in remote execution environments. When using the

deploy()

function in Python, you can configure this behavior programmatically. The

pull

section consists of a series of steps that are executed in sequence, each with optional fields like

requires

and

id

. These fields map directly to Python keyword arguments when using the

deploy()

function. To define this in Python, you would typically set up your deployment with the necessary steps in a similar sequence. Unfortunately, the exact Python equivalent isn't directly documented, but you can refer to the

deploy

function in the Prefect API to set up your deployment programmatically. For more details, you can review the Prefect 3.x documentation on deployment actions. If you have specific steps or configurations in mind, you might need to translate them into Python code that mimics the behavior of the steps in the

prefect.yaml

file.

.from_source

is the short answer

@Marvin @Nate But how would I add additional steps like e.g.

prefect.deployments.steps.set_working_directory

or

prefect.deployments.steps.pip_install_requirements

?

To add additional steps like

set_working_directory

or

pip_install_requirements

in your Python deployment script, you can use the functions provided in the

prefect.deployments.steps

module. Here's how you can incorporate these steps: 1. Set Working Directory: Use the

set_working_directory

function from the

prefect.deployments.steps.pull

module. 2. Pip Install Requirements: Use the

pip_install_requirements

function from the

prefect.deployments.steps.utility

module. Here's a basic example of how you might set these up in a Python script:

python
from prefect.deployments.steps.pull import set_working_directory
from prefect.deployments.steps.utility import pip_install_requirements

# Set the working directory
set_working_directory('/path/to/working/directory')

# Install requirements
pip_install_requirements('/path/to/requirements.txt')

Ensure that you adjust the paths to match your specific requirements. These functions help in setting up your environment as part of the deployment process.

well you can always manually add

pull_steps

directly on the deployment object after you create it (not common), but essentially this is a tradeoff associated with using the python interface instead of yaml with the yaml its really easy to add arbitrary stuff like entire new steps like this (defined here)

I am currently using

prefect.yaml

with several steps. However I wonder whether I should migrate to python deployments instead. Main reason is that I plan to use `run_deployment()`in my flows. And by using python I can just define a constant for the deployment name. If the name is defined in

prefect.yaml

I need to duplicate the string.

you could use prefect variables, which you can template into your yaml and also read in python

Valid point. I'll consider that. Thanks a ton!