fix: Properly sample Array columns with nested inner types

[Copilot]

Sampling failed for Array(List(...), ...) and other nested array types because _sample_unchecked assumed inner.sample() returned flat scalars suitable for reshape(). When the inner type is a complex nested type, it returns structured objects that cannot be reshaped.

Changes

• dataframely/columns/array.py: Added type checking in _sample_unchecked to detect nested inner types (List, Array, Struct) and construct the array structure manually via a recursive helper function. Scalar inner types continue using the optimized reshape() path.

• tests/columns/test_sample.py: Added parametrized tests covering Array(List(...)), Array(Array(...)), and Array(Struct(...)) with various shapes and sample counts.

Example

import dataframely as dy

class TestSchema(dy.Schema):
    a = dy.Array(dy.List(dy.Bool()), 2)

# Previously: InvalidOperationError: cannot reshape array of size 49 into shape (1, 2)
df = TestSchema.sample(1, generator=dy.random.Generator(0))  # Now works

Original prompt

---

This section details on the original issue you should resolve

<issue_title>Sampling fails for columns of dtype Array(List(...), ...)</issue_title>
<issue_description>Consider:

def test_sample_array_list(arr_size, n_samples):
    class TestSchema(dy.Schema):
        a = dy.Array(dy.List(dy.Bool()), arr_size)
    TestSchema.sample(n_samples, generator=dy.random.Generator(0))

On dataframely==2.3.1:

• calling test_sample_array_list(2, 1)
  fails with polars.exceptions.InvalidOperationError: cannot reshape array of size 49 into shape (1, 2)
• and calling test_sample_array_list(1, 1)
  fails with polars.exceptions.SchemaError: type Boolean is incompatible with expected type List(Boolean)

I'd expect both to pass.</issue_description>

Comments on the Issue (you are @copilot in this section)

• Fixes Sampling fails for columns of dtype Array(List(...), ...) #243

---

💡 You can make Copilot smarter by setting up custom instructions, customizing its development environment and configuring Model Context Protocol (MCP) servers. Learn more Copilot coding agent tips in the docs.

[codecov]

Codecov Report

❌ Patch coverage is 66.66667% with 8 lines in your changes missing coverage. Please review.
✅ Project coverage is 99.74%. Comparing base (b813ac9) to head (c3007ee).

Files with missing lines	Patch %	Lines
dataframely/columns/array.py	66.66%	8 Missing ⚠️

Additional details and impacted files

@@             Coverage Diff             @@
##              main     #244      +/-   ##
===========================================
- Coverage   100.00%   99.74%   -0.26%     
===========================================
  Files           53       53              
  Lines         3066     3089      +23     
===========================================
+ Hits          3066     3081      +15     
- Misses           0        8       +8     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.