feat: add Series properties: values, shapes, ndim, size, and hasnans (#13)

* chore: add spec describing all pandas methods.

* docs: Add detailed steps to series methods spec (#10)

Updates `specs/2025-09-16-series-methods.md` to include a more
detailed set of instructions for implementing the series methods.

The new instructions guide the developer to:
- Read the specification file before starting.
- Mark implemented methods on the checklist.
- Reset the checklist before submitting a pull request.

Co-authored-by: google-labs-jules[bot] <161369871+google-labs-jules[bot]@users.noreply.github.com>

* Implement six new properties for the leanframe Series class. (#11)

* Implement six new properties for the leanframe Series class.

This change adds the following pandas-compatible properties to `leanframe.core.series.Series`:
- `values`
- `shape`
- `nbytes`
- `ndim`
- `size`
- `hasnans`

The implementations are based on the underlying Ibis and PyArrow objects to ensure reasonable performance.

Unit tests have been added to verify the correctness of these new properties, including handling of null (NaN) values.

The specification file `specs/2025-09-16-series-methods.md` has been updated to reflect these changes.

* refactor(tests): Split Series properties tests into separate functions (#12)

This commit refactors the tests for the `Series` properties in `tests/unit/test_series.py`.

The single `test_series_properties` function has been replaced with six separate test functions, one for each of the following properties:
- `ndim`
- `size`
- `shape`
- `hasnans`
- `values`
- `nbytes`

A pytest fixture `series_for_properties` has been introduced to provide the test data for these new functions, reducing code duplication and improving test isolation.

Co-authored-by: google-labs-jules[bot] <161369871+google-labs-jules[bot]@users.noreply.github.com>

* Apply suggestions from code review

---------

Co-authored-by: google-labs-jules[bot] <161369871+google-labs-jules[bot]@users.noreply.github.com>

* Update leanframe/core/series.py

---------

Co-authored-by: google-labs-jules[bot] <161369871+google-labs-jules[bot]@users.noreply.github.com>

Author: tswast

leanframe/core/series.py

@@ -17,6 +17,7 @@
 from __future__ import annotations
 
 import ibis.expr.types as ibis_types
+import numpy as np
 import pandas as pd
 
 from leanframe.core.dtypes import convert_ibis_to_pandas
@@ -41,6 +42,36 @@ def dtype(self) -> pd.ArrowDtype:
     def name(self) -> str:
         """Name of the column."""
         return self._data.get_name()
+
+    @property
+    def values(self) -> np.ndarray:
+        """Return a numpy representation of the Series."""
+        return self._data.to_pyarrow().to_numpy()
+
+    @property
+    def shape(self) -> tuple[int, ...]:
+        """Return a tuple of the shape of the underlying data."""
+        return (self.size,)
+
+    @property
+    def nbytes(self) -> int:
+        """Return the number of bytes in the underlying data."""
+        raise NotImplementedError("nbytes not relevant for ibis expression.")
+
+    @property
+    def ndim(self) -> int:
+        """Return the number of dimensions of the underlying data."""
+        return 1
+
+    @property
+    def size(self) -> int:
+        """Return the number of elements in the underlying data."""
+        return self._data.as_table().count().to_pyarrow().as_py()
+
+    @property
+    def hasnans(self) -> bool:
+        """Return True if there are any NaNs, False otherwise."""
+        return self._data.isnull().any().to_pyarrow().as_py()
 
     def __add__(self, other) -> Series:
         return Series(self._data + getattr(other, "_data", other))

specs/2025-09-16-series-methods.md

@@ -0,0 +1,276 @@
+# Series methods
+
+Implement scalar, aggregate, and window methods for leanframe Series objects.
+
+## Background
+
+The leanframe package aims to provide a pandas-compatible interface while
+maintaining a 1:1 mapping to related Ibis types. DataFrame is 1:1 with Ibis
+Table, and Series is 1:1 with Ibis Column.
+
+Pandas features not supported by Ibis, such as specifying index column(s) or a
+total ordering over all rows are also not supported by leanframe.
+
+## Acceptance Criteria
+
+Implement each pandas Series method that doesn't require an index or
+ordering. When an item is completed, edit `specs/2025-09-16-series-methods.md`
+and check off the item with an x.
+
+- [x] pandas.Series.index -- not feasible, requires index
+- [ ] pandas.Series.array
+- [x] pandas.Series.values
+- [x] pandas.Series.dtype
+- [x] pandas.Series.shape
+- [x] pandas.Series.nbytes
+- [x] pandas.Series.ndim
+- [x] pandas.Series.size
+- [x] pandas.Series.T -- not feasible
+- [ ] pandas.Series.memory_usage
+- [x] pandas.Series.hasnans
+- [ ] pandas.Series.empty
+- [ ] pandas.Series.dtypes
+- [x] pandas.Series.name
+- [ ] pandas.Series.flags
+- [ ] pandas.Series.set_flags
+- [ ] pandas.Series.astype
+- [ ] pandas.Series.convert_dtypes
+- [ ] pandas.Series.infer_objects
+- [ ] pandas.Series.copy
+- [ ] pandas.Series.bool
+- [ ] pandas.Series.to_numpy
+- [ ] pandas.Series.to_period
+- [ ] pandas.Series.to_timestamp
+- [ ] pandas.Series.to_list
+- [ ] pandas.Series.__array__
+- [ ] pandas.Series.get
+- [x] pandas.Series.at -- not feasible, requires index
+- [x] pandas.Series.iat -- not feasible, requires ordering
+- [x] pandas.Series.loc -- not feasible, requires index
+- [x] pandas.Series.iloc -- not feasible, requires ordering
+- [ ] pandas.Series.__iter__
+- [x] pandas.Series.items -- not feasible, requires index
+- [ ] pandas.Series.keys -- not feasible, requires index
+- [ ] pandas.Series.pop -- not feasible, requires index
+- [ ] pandas.Series.item
+- [ ] pandas.Series.xs
+- [ ] pandas.Series.add
+- [ ] pandas.Series.sub
+- [ ] pandas.Series.mul
+- [ ] pandas.Series.div
+- [ ] pandas.Series.truediv
+- [ ] pandas.Series.floordiv
+- [ ] pandas.Series.mod
+- [ ] pandas.Series.pow
+- [ ] pandas.Series.radd
+- [ ] pandas.Series.rsub
+- [ ] pandas.Series.rmul
+- [ ] pandas.Series.rdiv
+- [ ] pandas.Series.rtruediv
+- [ ] pandas.Series.rfloordiv
+- [ ] pandas.Series.rmod
+- [ ] pandas.Series.rpow
+- [ ] pandas.Series.combine
+- [ ] pandas.Series.combine_first
+- [ ] pandas.Series.round
+- [ ] pandas.Series.lt
+- [ ] pandas.Series.gt
+- [ ] pandas.Series.le
+- [ ] pandas.Series.ge
+- [ ] pandas.Series.ne
+- [ ] pandas.Series.eq
+- [ ] pandas.Series.product
+- [ ] pandas.Series.dot
+- [ ] pandas.Series.apply
+- [ ] pandas.Series.agg
+- [ ] pandas.Series.aggregate
+- [ ] pandas.Series.transform
+- [ ] pandas.Series.map
+- [ ] pandas.Series.groupby
+- [ ] pandas.Series.rolling
+- [ ] pandas.Series.expanding
+- [ ] pandas.Series.ewm
+- [ ] pandas.Series.pipe
+- [ ] pandas.Series.abs
+- [ ] pandas.Series.all
+- [ ] pandas.Series.any
+- [ ] pandas.Series.autocorr
+- [ ] pandas.Series.between
+- [ ] pandas.Series.clip
+- [ ] pandas.Series.corr
+- [ ] pandas.Series.count
+- [ ] pandas.Series.cov
+- [ ] pandas.Series.cummax
+- [ ] pandas.Series.cummin
+- [ ] pandas.Series.cumprod
+- [ ] pandas.Series.cumsum
+- [ ] pandas.Series.describe
+- [ ] pandas.Series.diff
+- [ ] pandas.Series.factorize
+- [ ] pandas.Series.kurt
+- [ ] pandas.Series.max
+- [ ] pandas.Series.mean
+- [ ] pandas.Series.median
+- [ ] pandas.Series.min
+- [ ] pandas.Series.mode
+- [ ] pandas.Series.nlargest
+- [ ] pandas.Series.nsmallest
+- [ ] pandas.Series.pct_change
+- [ ] pandas.Series.prod
+- [ ] pandas.Series.quantile
+- [ ] pandas.Series.rank
+- [ ] pandas.Series.sem
+- [ ] pandas.Series.skew
+- [ ] pandas.Series.std
+- [ ] pandas.Series.sum
+- [ ] pandas.Series.var
+- [ ] pandas.Series.kurtosis
+- [ ] pandas.Series.unique
+- [ ] pandas.Series.nunique
+- [ ] pandas.Series.is_unique
+- [ ] pandas.Series.is_monotonic_increasing
+- [ ] pandas.Series.is_monotonic_decreasing
+- [ ] pandas.Series.value_counts
+- [ ] pandas.Series.align
+- [ ] pandas.Series.case_when
+- [ ] pandas.Series.drop
+- [ ] pandas.Series.droplevel
+- [ ] pandas.Series.drop_duplicates
+- [ ] pandas.Series.duplicated
+- [ ] pandas.Series.equals
+- [ ] pandas.Series.first
+- [ ] pandas.Series.head
+- [ ] pandas.Series.idxmax
+- [ ] pandas.Series.idxmin
+- [ ] pandas.Series.isin
+- [ ] pandas.Series.last
+- [ ] pandas.Series.reindex
+- [ ] pandas.Series.reindex_like
+- [ ] pandas.Series.rename
+- [ ] pandas.Series.rename_axis
+- [ ] pandas.Series.reset_index
+- [ ] pandas.Series.sample
+- [ ] pandas.Series.set_axis
+- [ ] pandas.Series.take
+- [ ] pandas.Series.tail
+- [ ] pandas.Series.truncate
+- [ ] pandas.Series.where
+- [ ] pandas.Series.mask
+- [ ] pandas.Series.add_prefix
+- [ ] pandas.Series.add_suffix
+- [ ] pandas.Series.filter
+- [ ] pandas.Series.backfill
+- [ ] pandas.Series.bfill
+- [ ] pandas.Series.dropna
+- [ ] pandas.Series.ffill
+- [ ] pandas.Series.fillna
+- [ ] pandas.Series.interpolate
+- [ ] pandas.Series.isna
+- [ ] pandas.Series.isnull
+- [ ] pandas.Series.notna
+- [ ] pandas.Series.notnull
+- [ ] pandas.Series.pad
+- [ ] pandas.Series.replace
+- [ ] pandas.Series.argsort
+- [ ] pandas.Series.argmin
+- [ ] pandas.Series.argmax
+- [ ] pandas.Series.reorder_levels
+- [ ] pandas.Series.sort_values
+- [ ] pandas.Series.sort_index
+- [ ] pandas.Series.swaplevel
+- [ ] pandas.Series.unstack
+- [ ] pandas.Series.explode
+- [ ] pandas.Series.searchsorted
+- [ ] pandas.Series.ravel
+- [ ] pandas.Series.repeat
+- [ ] pandas.Series.squeeze
+- [ ] pandas.Series.view
+- [ ] pandas.Series.compare
+- [ ] pandas.Series.update
+- [ ] pandas.Series.asfreq
+- [ ] pandas.Series.asof
+- [ ] pandas.Series.shift
+- [ ] pandas.Series.first_valid_index
+- [ ] pandas.Series.last_valid_index
+- [ ] pandas.Series.resample
+- [ ] pandas.Series.tz_convert
+- [ ] pandas.Series.tz_localize
+- [ ] pandas.Series.at_time
+- [ ] pandas.Series.between_time
+- [ ] pandas.Series.str
+- [ ] pandas.Series.cat
+- [ ] pandas.Series.dt
+- [ ] pandas.Series.sparse
+- [ ] pandas.Series.attrs
+- [ ] pandas.Series.hist
+- [ ] pandas.Series.to_pickle
+- [ ] pandas.Series.to_csv
+- [ ] pandas.Series.to_dict
+- [ ] pandas.Series.to_excel
+- [ ] pandas.Series.to_frame
+- [ ] pandas.Series.to_xarray
+- [ ] pandas.Series.to_hdf
+- [ ] pandas.Series.to_sql
+- [ ] pandas.Series.to_json
+- [ ] pandas.Series.to_string
+- [ ] pandas.Series.to_clipboard
+- [ ] pandas.Series.to_latex
+- [ ] pandas.Series.to_markdown
+
+## Detailed Steps
+
+This document outlines the steps to implement new Series methods. Follow these
+steps carefully.
+
+### 1. Understand the Task
+
+- [ ] **Read this document carefully**: Before you begin, read this entire
+      document (`specs/2025-09-16-series-methods.md`) to understand the scope
+      and requirements of the task.
+
+### 2. Implement a Method
+
+- [ ] **Choose a method**: Select an unchecked method from the `Acceptance
+      Criteria` list above.
+- [ ] **Feasibility check**: Determine if the method is feasible to implement
+      given the constraints mentioned in the `Background` section.
+- [ ] **Mark as complete or infeasible**:
+    - If the method is **feasible**, continue to the next step.
+    - If the method is **not feasible**, mark it with an `x` in the `Acceptance
+      Criteria` list and add a brief note explaining why (e.g., `- [x]
+      pandas.Series.index -- not feasible, requires index`).
+- [ ] **Implement the method**: Add the method to `leanframe/core/series.py`.
+      Ensure your implementation is consistent with the existing codebase.
+- [ ] **Add unit tests**: Create comprehensive unit tests for the new method in
+      `tests/unit/test_series.py`. Cover edge cases and different data types.
+
+### 3. Verify Your Changes
+
+- [ ] **Run tests**: Execute all tests by running `uv run pytest tests` to
+      ensure your changes haven't introduced any regressions.
+- [ ] **Run static analysis**: Run `uv run mypy leanframe tests` and `uv run
+      ruff check` to check for type errors and linting issues.
+- [ ] **Mark the method as complete**: Once the implementation is complete and
+      all checks pass, edit this file (`specs/2025-09-16-series-methods.md`) and
+      mark the method you implemented with an `x` in the `Acceptance Criteria`
+      list.
+
+### 4. Finalizing for Submission
+
+- [ ] **Reset checkboxes**: Before submitting your pull request, uncheck all the
+      boxes in the `Acceptance Criteria` that you have marked with an `x` during
+      your work. The spec file should be in a clean state for the next
+      developer. Leave the originally checked items as they are.
+
+## Verification
+
+*Specify the commands to run to verify the changes.*
+
+- [ ] All new and existing tests `uv run pytest tests` should pass.
+- [ ] The `uv run mypy leanframe tests` static type checker should pass.
+- [ ] The `uv run ruff check` linter should pass.
+- [ ] Only add git commits. Do not change git history.
+
+## Constraints
+
+Follow the guidelines listed in GEMINI.md at the root of the repository.

tests/unit/test_series.py

@@ -19,9 +19,61 @@
 import pyarrow as pa
 import pytest
 
+import numpy as np
+
 import leanframe
 
 
+@pytest.fixture
+def series_for_properties(session):
+    df_pd = pd.DataFrame(
+        {
+            "int_col": pd.Series([1, 2, 3], dtype=pd.ArrowDtype(pa.int64())),
+            "float_col": pd.Series(
+                [1.0, float("nan"), 3.0], dtype=pd.ArrowDtype(pa.float64())
+            ),
+        }
+    )
+    df_lf = session.DataFrame(df_pd)
+    return df_lf["int_col"], df_lf["float_col"]
+
+
+def test_series_ndim(series_for_properties):
+    series_int, series_float = series_for_properties
+    assert series_int.ndim == 1
+    assert series_float.ndim == 1
+
+
+def test_series_size(series_for_properties):
+    series_int, series_float = series_for_properties
+    assert series_int.size == 3
+    assert series_float.size == 3
+
+
+def test_series_shape(series_for_properties):
+    series_int, series_float = series_for_properties
+    assert series_int.shape == (3,)
+    assert series_float.shape == (3,)
+
+
+def test_series_hasnans(series_for_properties):
+    series_int, series_float = series_for_properties
+    assert not series_int.hasnans
+    assert series_float.hasnans
+
+
+def test_series_values(series_for_properties):
+    series_int, series_float = series_for_properties
+    np.testing.assert_array_equal(series_int.values, np.array([1, 2, 3]))
+
+
+def test_series_nbytes(series_for_properties):
+    series_int, series_float = series_for_properties
+    
+    with pytest.raises(NotImplementedError, match="nbytes"):
+        assert series_int.nbytes
+
+
 @pytest.mark.parametrize(
     ("column", "expected_dtype"),
     [