Exam DP-100 topic 3 question 6 discussion

it should be D

Correct answer is D. The dependencies of steps is implicitly defined in the step itself (via arguments parameter that could be an output from some other step). Read https://docs.microsoft.com/en-us/python/api/azureml-pipeline-core/azureml.pipeline.core.pipelinedata?view=azure-ml-py.

I would go for D

I go for D, chaudha4 explanation below is spot-on.

A would be correct if it had been p = Pipeline(ws, steps=StepSequence(steps=[initial_steps, step4])) https://learn.microsoft.com/en-us/python/api/azureml-pipeline-core/azureml.pipeline.core.stepsequence?view=azure-ml-py#remarks

It should be C Check these links: 1) https://learn.microsoft.com/en-us/python/api/azureml-pipeline-core/azureml.pipeline.core.builder.pipelinestep?view=azure-ml-py#azureml-pipeline-core-builder-pipelinestep-run-after 2) https://github.com/Azure/MachineLearningNotebooks/blob/master/how-to-use-azureml/machine-learning-pipelines/intro-to-pipelines/aml-pipelines-getting-started.ipynb -> section "Running a few steps in sequence"

C could be

A. is also wrong because it creates a non documented dependency of step_2_b on step_2_a

I will definitely go with D

it should be C

I don't want to create confusion but I was unable to find documentation about this, so I used chatGCP and it provides code samples, it seems that A is not an option cause the steps parameter should be a list of steps, not a nested list of steps. Option C the pipeline definition is missing the steps that step3 depends on, so like chaudha4 explains the dependencies are implicitly and I would go for option D

A gets my vote. D has absolutely no sequence and C fails to run step 2_a.

see the graph, so it's A

A is correct, that is the syntax I used for years!

D misses the StepSequence and it means it runs all the steps in parallel. Then C seems to be a better option. The only problem with C is that it doesn't run 1_a and 1_b in parallel and it runs them sequentially.

it should be C Running a few steps in sequence https://github.com/Azure/MachineLearningNotebooks/blob/master/how-to-use-azureml/machine-learning-pipelines/intro-to-pipelines/aml-pipelines-getting-started.ipynb

I think the answer is D I think the key here is this sentence: "but will wait for modules with dependencies." This means the lines indicate data dependencies between steps not just the sequence. run_after() or any such constructs are for sequencing where there are no data dependencies between the steps. Text from the notebook linked below: We will reuse step1, step2, step3, but build the pipeline in such a way that we chain step3 after step2 and step2 after step1. Note that there is no explicit data dependency between these steps, but still steps can be made dependent by using the run_after construct. https://github.com/Azure/MachineLearningNotebooks/blob/master/how-to-use-azureml/machine-learning-pipelines/intro-to-pipelines/aml-pipelines-getting-started.ipynb