# Advanced Topics


# Custom Pipeline Stages

Build complex analysis workflows by chaining stages:

``` python
from gdtest_nav_icons import Pipeline, analyze, summarize

pipe = Pipeline("full-analysis")
pipe.add_stage(lambda d: [x for x in d if x > 0])  # filter
pipe.add_stage(lambda d: [x ** 0.5 for x in d])    # sqrt

clean_data = pipe.run([-1, 4, 9, -2, 16, 25])
print(analyze(clean_data))
```


# Exporting Results

``` python
from gdtest_nav_icons import export_json

results = [{"metric": "accuracy", "value": 0.95}]
export_json(results, "output.json", indent=4)
```


# Summary Reports

``` python
from gdtest_nav_icons import summarize

print(summarize([88, 92, 79, 95, 100], label="Exam scores"))
```