A reactive notebook for Python — run reproducible experiments, query with SQL, execute as a script, deploy as an app, and version with git. Stored as pure Python. All in a modern, AI-native editor.
Python artificial-intelligencedagdata-sciencedata-visualizationdataflowdeveloper-toolsmachine-learningnotebookspipelinepythonreactivesqlweb-app
from __future__ import annotations
from typing import Any
import narwhals.stable.v2 as nw
import pytest
from marimo._dependencies.dependencies import DependencyManager
from marimo._plugins.ui._impl.table import SortArgs
from marimo._plugins.ui._impl.tables.narwhals_table import NarwhalsTableManager
from marimo._plugins.ui._impl.tables.selection import (
INDEX_COLUMN_NAME,
add_selection_column,
remove_selection_column,
)
try:
import pandas as pd
import polars as pl
import pyarrow as pa
except ImportError:
pl = None
pd = None
pa = None
BACKENDS = [pl, pd, pa]
HAS_DEPS = (
DependencyManager.polars.has()
and DependencyManager.pandas.has()
and DependencyManager.pyarrow.has()
)
@pytest.mark.skipif(not HAS_DEPS, reason="Deps not installed")
@pytest.mark.parametrize("backend", BACKENDS)
def test_selection_with_index_column(backend: Any):
# Create test data with index column
data = nw.from_dict(
{
INDEX_COLUMN_NAME: [0, 1, 2],
"name": ["Alice", "Bob", "Charlie"],
"age": [30, 25, 35],
},
backend=backend,
)
manager = NarwhalsTableManager(data)
# Test selection using index column
selected = manager.select_rows([0, 2])
result = selected.data.to_dict(as_series=False)
assert result[INDEX_COLUMN_NAME] == [0, 2]
assert result["name"] == ["Alice", "Charlie"]
assert result["age"] == [30, 35]
@pytest.mark.skipif(not HAS_DEPS, reason="Deps not installed")
@pytest.mark.parametrize("backend", BACKENDS)
def test_selection_with_index_column_and_sort(backend: Any):
# Create test data with index column
data = nw.from_dict(
{
INDEX_COLUMN_NAME: [0, 1, 2],
"name": ["Alice", "Bob", "Charlie"],
"age": [30, 25, 35],
},
backend=backend,
)
manager = NarwhalsTableManager(data)
# Sort and select
sorted_data = manager.sort_values([SortArgs(by="age", descending=True)])
selected = sorted_data.select_rows([0, 2])
result = selected.data.to_dict(as_series=False)
assert result[INDEX_COLUMN_NAME] == [2, 0] # Original indices preserved
assert result["name"] == ["Charlie", "Alice"]
assert result["age"] == [35, 30]
@pytest.mark.skipif(not HAS_DEPS, reason="Deps not installed")
@pytest.mark.parametrize("backend", BACKENDS)
def test_selection_with_index_column_and_search(backend: Any):
# Create test data with index column
data = nw.from_dict(
{
INDEX_COLUMN_NAME: [0, 1, 2],
"name": ["Alice", "Bob", "Charlie"],
"age": [30, 25, 35],
},
backend=backend,
)
manager = NarwhalsTableManager(data)
# Search and select
searched = manager.search("ali")
selected = searched.select_rows([0])
result = selected.data.to_dict(as_series=False)
assert result[INDEX_COLUMN_NAME] == [0]
assert result["name"] == ["Alice"]
assert result["age"] == [30]
@pytest.mark.skipif(not HAS_DEPS, reason="Deps not installed")
@pytest.mark.parametrize("backend", BACKENDS)
def test_selection_with_index_column_empty(backend: Any):
# Create test data with index column
data = nw.from_dict(
{INDEX_COLUMN_NAME: [0, 1], "name": ["Alice", "Bob"], "age": [30, 25]},
backend=backend,
)
manager = NarwhalsTableManager(data)
# Test empty selection
selected = manager.select_rows([])
result = selected.data.to_dict(as_series=False)
assert result[INDEX_COLUMN_NAME] == []
assert result["name"] == []
assert result["age"] == []
@pytest.mark.skipif(not HAS_DEPS, reason="Deps not installed")
@pytest.mark.parametrize("backend", BACKENDS)
def test_selection_without_index_column(backend: Any):
# Create test data without index column
data = nw.from_dict(
{"name": ["Alice", "Bob", "Charlie"], "age": [30, 25, 35]},
backend=backend,
)
manager = NarwhalsTableManager(data)
# Test selection falls back to positional indexing
selected = manager.select_rows([0, 2])
result = selected.data.to_dict(as_series=False)
assert "name" in result
assert result["name"] == ["Alice", "Charlie"]
assert result["age"] == [30, 35]
@pytest.mark.skipif(not HAS_DEPS, reason="Deps not installed")
@pytest.mark.parametrize("backend", BACKENDS)
def test_selection_with_index_column_take(backend: Any):
# Create test data with index column
data = nw.from_dict(
{
INDEX_COLUMN_NAME: [0, 1, 2, 3, 4],
"name": ["Alice", "Bob", "Charlie", "Dave", "Eve"],
"age": [30, 25, 35, 28, 22],
},
backend=backend,
)
manager = NarwhalsTableManager(data)
# Test take operation preserves index column
taken = manager.take(2, 1) # Take 2 rows starting from index 1
result = taken.data.to_dict(as_series=False)
assert result[INDEX_COLUMN_NAME] == [1, 2]
assert result["name"] == ["Bob", "Charlie"]
assert result["age"] == [25, 35]
@pytest.mark.skipif(not HAS_DEPS, reason="Deps not installed")
@pytest.mark.parametrize("backend", BACKENDS)
def test_add_selection_to_dataframe(backend: Any):
data = nw.from_dict(
{"name": ["Alice", "Bob", "Charlie"], "age": [30, 25, 35]},
backend=backend,
)
with_selection, has_stable_row_id = add_selection_column(data.to_native())
assert has_stable_row_id is True
# Convert back to Narwhals to assert
nw_df = nw.from_native(with_selection)
assert nw_df.columns == [INDEX_COLUMN_NAME, "name", "age"]
assert nw_df[INDEX_COLUMN_NAME].to_list() == [0, 1, 2]
@pytest.mark.skipif(not HAS_DEPS, reason="Deps not installed")
@pytest.mark.parametrize("backend", BACKENDS)
def test_add_selection_to_dataframe_already_has_index(backend: Any):
data = nw.from_dict(
{
INDEX_COLUMN_NAME: [0, 1, 2],
"name": ["Alice", "Bob", "Charlie"],
"age": [30, 25, 35],
},
backend=backend,
)
with_selection, has_stable_row_id = add_selection_column(data.to_native())
assert has_stable_row_id is True
# Convert back to Narwhals to assert
nw_df = nw.from_native(with_selection)
assert nw_df.columns == [INDEX_COLUMN_NAME, "name", "age"]
assert nw_df[INDEX_COLUMN_NAME].to_list() == [0, 1, 2]
@pytest.mark.skipif(not HAS_DEPS, reason="Deps not installed")
@pytest.mark.parametrize("backend", BACKENDS)
def test_remove_selection_column(backend: Any):
data = nw.from_dict(
{"name": ["Alice", "Bob", "Charlie"], "age": [30, 25, 35]},
backend=backend,
)
with_selection, has_stable_row_id = add_selection_column(data.to_native())
assert has_stable_row_id is True
without_selection = remove_selection_column(with_selection)
nw_df = nw.from_native(without_selection)
assert nw_df.columns == ["name", "age"]
# Remove again
without_selection = remove_selection_column(without_selection)
nw_df = nw.from_native(without_selection)
assert nw_df.columns == ["name", "age"]