Fix Example 1 documentation: pipeline() returns numpy array, not dict

[Tao-Yida]

Example 1 in the documentation (https://augraphy.readthedocs.io/en/latest/doc/source/example_usage.html) contains incorrect code that causes an IndexError when users try to run it.
The current code shows:

data = pipeline(image)
augmented = data["output"]  # This causes IndexError!

ruturns:

IndexError: only integers, slices (`:`), ellipsis (`...`), numpy.newaxis (`None`) and integer or boolean arrays are valid indices

But pipeline(image) returns a numpy array directly (via __call__ method which calls augment(image, return_dict=0)), not a dictionary.

• Solution

1. Fixed the example to use the return value correctly:

augmented = pipeline(image)  # Returns numpy array directly

2. Explicitly request dictionary:

data = pipeline.augment(image, return_dict=1)
augmented = data["output"]

• Technical Details

In AugraphyPipeline.__call__:

def __call__(self, image):
    return self.augment(image, return_dict=0)  # return_dict=0 returns numpy array

When return_dict=0, the augment method returns data["output"] (numpy array), not the full dictionary.

[Copilot]

Pull request overview

This PR fixes incorrect documentation in Example 1 that would cause an IndexError when users run the code. The documentation incorrectly showed that pipeline(image) returns a dictionary, when it actually returns a numpy array directly.

Changes:

• Updated Example 1 to show two correct ways of using the pipeline: calling pipeline(image) to get a numpy array directly, and calling pipeline.augment(image, return_dict=1) to explicitly request a dictionary

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.