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
import ast
from datetime import datetime
from typing import Any
from unittest.mock import patch
import pytest
from marimo._data.charts import (
REPLACE_DF_COMMENT,
ChartBuilder,
DateChartBuilder,
get_chart_builder,
)
from marimo._data.models import DataType
from marimo._dependencies.dependencies import DependencyManager
from tests._data.mocks import NON_EAGER_LIBS, create_dataframes
from tests.mocks import snapshotter
TYPES: list[tuple[DataType, bool]] = [
("boolean", False),
("date", False),
("datetime", False),
("time", False),
("integer", False),
("number", False),
("string", False),
("string", True),
("unknown", False),
]
snapshot = snapshotter(__file__)
HAS_DEPS = (
DependencyManager.pandas.has()
and DependencyManager.altair.has()
and DependencyManager.polars.has()
)
@pytest.fixture
def mock_light_theme():
"""Mock the theme to always return 'light' for consistent test results."""
with patch("marimo._data.charts.get_current_theme", return_value="light"):
yield
def test_get_chart_builder():
for t, should_limit_to_10_items in TYPES:
assert isinstance(
get_chart_builder(t, should_limit_to_10_items), ChartBuilder
)
def test_geometry_uses_unknown_chart_builder() -> None:
geometry_code = get_chart_builder("geometry").altair_code("df", "geometry")
unknown_code = get_chart_builder("unknown").altair_code("df", "geometry")
assert geometry_code == unknown_code
def validate_and_return_codes(simple: bool) -> list[str]:
outputs: list[str] = []
for t, should_limit_to_10_items in TYPES:
builder = get_chart_builder(t, should_limit_to_10_items)
code = builder.altair_code("df", "some_column", simple=simple)
# Validate it is valid Python code
try:
ast.parse(code)
except SyntaxError as e:
raise SyntaxError(f"Invalid Python code for {t}") from e
title = f"{t} (limit to 10 items)" if should_limit_to_10_items else t
outputs.append(f"# {title}\n{code}")
return outputs
@pytest.mark.usefixtures("mock_light_theme")
def test_charts_altair_code():
outputs = validate_and_return_codes(simple=True)
snapshot("charts.txt", "\n\n".join(outputs))
complex_outputs = validate_and_return_codes(simple=False)
snapshot("charts-complex.txt", "\n\n".join(complex_outputs))
def test_charts_bad_characters():
outputs: list[str] = []
builder = get_chart_builder("string", False)
chars = {
"<": "angles",
"[0]": "brackets",
":": "colon",
".": "period",
"\\": "backslash",
}
for char, name in chars.items():
col = f"col{char}{name}"
code = builder.altair_code("df", col)
outputs.append(f"# {col}\n{code}")
snapshot("charts_bad_characters.txt", "\n\n".join(outputs))
def test_charts_altair_code_with_comment():
builder = get_chart_builder("string", False)
code = builder.altair_code_with_comment("df", "some_column", simple=True)
assert f"alt.Chart([]) {REPLACE_DF_COMMENT}" in code
@pytest.mark.skipif(not HAS_DEPS, reason="optional dependencies not installed")
@pytest.mark.usefixtures("mock_light_theme")
def test_charts_altair_json():
outputs: list[str] = []
import altair as alt
import pandas as pd
data = pd.DataFrame({"some_column": [1, 2, 3]})
for t, should_limit_to_10_items in TYPES:
builder = get_chart_builder(t, should_limit_to_10_items)
code = builder.altair_json(data, "some_column")
# Validate it is valid JSON
alt.Chart.from_json(code)
title = f"{t} (limit to 10 items)" if should_limit_to_10_items else t
outputs.append(f"# {title}\n{code}")
snapshot("charts_json.txt", "\n\n".join(outputs))
@pytest.mark.skipif(not HAS_DEPS, reason="optional dependencies not installed")
@pytest.mark.usefixtures("mock_light_theme")
def test_charts_altair_json_bad_data():
import altair as alt
import pandas as pd
data = pd.DataFrame({"some[0]really.bad:column": [1, 2, 3]})
build = get_chart_builder("string", False)
code = build.altair_json(data, "some[0]really.bad:column")
# Validate it is valid JSON
alt.Chart.from_json(code)
snapshot("charts_json_bad_data.txt", code)
@pytest.mark.skipif(not HAS_DEPS, reason="optional dependencies not installed")
@pytest.mark.parametrize(
("data", "expected_format"),
[
# More than 10 years
(
({"dates": [datetime(2000, 1, 1), datetime(2020, 1, 1)]}),
("%Y", "year"),
),
# More than a year
(
({"dates": [datetime(2020, 1, 1), datetime(2021, 6, 1)]}),
("%Y-%m", "yearmonth"),
),
# More than a month
(
({"dates": [datetime(2020, 1, 1), datetime(2020, 2, 15)]}),
("%Y-%m-%d", "yearmonthdate"),
),
# More than a day
(
(
{
"dates": [
datetime(2020, 1, 1, 0, 0),
datetime(2020, 1, 2, 12, 0),
]
}
),
("%Y-%m-%d %H:%M", "yearmonthdatehoursminutes"),
),
# Less than a day
(
(
{
"dates": [
datetime(2020, 1, 1, 0, 0, 0),
datetime(2020, 1, 1, 12, 30, 45),
]
}
),
("%Y-%m-%d %H:%M", "yearmonthdatehoursminutes"),
),
],
)
def test_date_chart_builder_guess_date_format_with_dataframes(
data: dict[str, list[Any]], expected_format: str
):
eager_dfs = create_dataframes(data, exclude=NON_EAGER_LIBS)
assert len(eager_dfs) > 0
builder = DateChartBuilder()
for df in eager_dfs:
date_format, time_unit = builder._guess_date_format(df, "dates")
assert (
date_format,
time_unit,
) == expected_format, f"Expected {expected_format} for {type(df)}"
non_eager_dfs = create_dataframes(data, include=NON_EAGER_LIBS)
for df in non_eager_dfs:
date_format, time_unit = builder._guess_date_format(df, "dates")
assert (
date_format,
time_unit,
) == (
DateChartBuilder.DEFAULT_DATE_FORMAT,
DateChartBuilder.DEFAULT_TIME_UNIT,
), f"Expected {DateChartBuilder.DEFAULT_DATE_FORMAT} for {type(df)}"
@pytest.mark.skipif(not HAS_DEPS, reason="optional dependencies not installed")
def test_date_chart_builder_guess_date_format_with_non_narwhalifiable_data():
builder = DateChartBuilder()
# Test with non-narwhalifiable data
class NonNarwhalifiableData:
pass
date_format, time_unit = builder._guess_date_format(
NonNarwhalifiableData(), "dates"
)
assert date_format == DateChartBuilder.DEFAULT_DATE_FORMAT
assert time_unit == DateChartBuilder.DEFAULT_TIME_UNIT
@pytest.mark.skipif(not HAS_DEPS, reason="optional dependencies not installed")
def test_date_chart_builder_get_date_format():
from datetime import date, datetime, time
import pandas as pd
import polars as pl
builder = DateChartBuilder()
data = pd.DataFrame(
{"dates": [datetime(2020, 1, 1), datetime(2020, 2, 1)]}
)
date_format, time_unit = builder._guess_date_format(data, "dates")
assert date_format == "%Y-%m-%d %H"
assert time_unit == "yearmonthdatehours"
# Test with Polars time types
data = pl.DataFrame(
{
"dates": pl.Series("dates", [date(2021, 1, 1)], dtype=pl.Date),
"times": pl.Series("times", [time(12, 0, 0)], dtype=pl.Time),
"datetimes": pl.Series(
"datetimes", [datetime.now()], dtype=pl.Datetime
),
},
)
date_format, time_unit = builder._guess_date_format(data, "dates")
assert date_format == "%Y-%m-%d %H:%M"
assert time_unit == "yearmonthdatehoursminutes"
date_format, time_unit = builder._guess_date_format(data, "times")
assert date_format == "%H:%M:%S"
assert time_unit == "hoursminutesseconds"
date_format, time_unit = builder._guess_date_format(data, "datetimes")
assert date_format == "%Y-%m-%d %H:%M"
assert time_unit == "yearmonthdatehoursminutes"