Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Fix hist #474

Merged
merged 3 commits into from
Dec 7, 2022
Merged

Fix hist #474

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
605ca0d
fix metric
Dec 7, 2022
2095ed9
Linters
Dec 7, 2022
70ea002
Linters for tests
Dec 7, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
67 changes: 24 additions & 43 deletions src/evidently/metrics/data_quality/column_distribution_metric.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,18 +12,22 @@
from evidently.renderers.base_renderer import MetricRenderer
from evidently.renderers.base_renderer import default_renderer
from evidently.renderers.html_widgets import HistogramData
from evidently.renderers.html_widgets import WidgetSize
from evidently.renderers.html_widgets import header_text
from evidently.renderers.html_widgets import histogram
from evidently.renderers.html_widgets import plotly_figure
from evidently.renderers.render_utils import get_distribution_plot_figure
from evidently.utils.data_operations import process_columns
from evidently.utils.data_operations import recognize_column_type
from evidently.utils.types import ColumnDistribution
from evidently.utils.visualizations import Distribution
from evidently.utils.visualizations import get_distribution_for_column


@dataclasses.dataclass
class ColumnDistributionMetricResult:
column_name: str
current: ColumnDistribution
reference: Optional[ColumnDistribution] = None
current: Distribution
reference: Optional[Distribution] = None


class ColumnDistributionMetric(Metric[ColumnDistributionMetricResult]):
Expand All @@ -37,14 +41,6 @@ def __init__(
) -> None:
self.column_name = column_name

@staticmethod
def _calculate_distribution(
column_name: str, dataset: pd.DataFrame, column_mapping: ColumnMapping
) -> ColumnDistribution:
columns = process_columns(dataset, column_mapping)
column_type = recognize_column_type(dataset=dataset, column_name=column_name, columns=columns)
return calculate_column_distribution(dataset[column_name], column_type)

def calculate(self, data: InputData) -> ColumnDistributionMetricResult:
if self.column_name not in data.current_data:
raise ValueError(f"Column '{self.column_name}' was not found in current data.")
Expand All @@ -53,15 +49,17 @@ def calculate(self, data: InputData) -> ColumnDistributionMetricResult:
if self.column_name not in data.reference_data:
raise ValueError(f"Column '{self.column_name}' was not found in reference data.")

current = self._calculate_distribution(self.column_name, data.current_data, data.column_mapping)

columns = process_columns(data.current_data, data.column_mapping)
column_type = recognize_column_type(dataset=data.current_data, column_name=self.column_name, columns=columns)
current_column = data.current_data[self.column_name]
reference_column = None
if data.reference_data is not None:
reference: Optional[ColumnDistribution] = self._calculate_distribution(
self.column_name, data.reference_data, data.column_mapping
)

else:
reference = None
reference_column = data.reference_data[self.column_name]
current, reference = get_distribution_for_column(
column_type=column_type,
current=current_column,
reference=reference_column,
)

return ColumnDistributionMetricResult(
column_name=self.column_name,
Expand All @@ -74,37 +72,20 @@ def calculate(self, data: InputData) -> ColumnDistributionMetricResult:
class ColumnDistributionMetricRenderer(MetricRenderer):
def render_json(self, obj: ColumnDistributionMetric) -> dict:
result = dataclasses.asdict(obj.get_result())
result.pop("current_distribution", None)
result.pop("reference_distribution", None)
result.pop("current", None)
result.pop("reference", None)
return result

def render_html(self, obj: ColumnDistributionMetric) -> List[BaseWidgetInfo]:
metric_result = obj.get_result()
current_histogram = HistogramData(
name="current",
x=list(metric_result.current.keys()),
y=list(metric_result.current.values()),
distr_fig = get_distribution_plot_figure(
current_distribution=metric_result.current,
reference_distribution=metric_result.reference,
color_options=self.color_options,
)

if metric_result.reference is not None:
reference_histogram: Optional[HistogramData] = HistogramData(
name="reference",
x=list(metric_result.reference.keys()),
y=list(metric_result.reference.values()),
)

else:
reference_histogram = None

result = [
header_text(label=f"Distribution for column '{metric_result.column_name}'."),
histogram(
title="",
primary_hist=current_histogram,
secondary_hist=reference_histogram,
color_options=self.color_options,
xaxis_title=metric_result.column_name,
yaxis_title="Count",
),
plotly_figure(title="", figure=distr_fig, size=WidgetSize.FULL),
]
return result
10 changes: 6 additions & 4 deletions tests/metrics/data_quality/test_column_distribution_metric.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,6 +8,7 @@
from evidently.metrics.data_quality.column_distribution_metric import ColumnDistributionMetric
from evidently.metrics.data_quality.column_distribution_metric import ColumnDistributionMetricResult
from evidently.report import Report
from evidently.utils.visualizations import Distribution


@pytest.mark.parametrize(
Expand All @@ -19,7 +20,7 @@
ColumnDistributionMetric(column_name="category_feature"),
ColumnDistributionMetricResult(
column_name="category_feature",
current={"n": 3, "d": 2, "p": 1},
current=Distribution(x=pd.Series(["n", "d", "p"]), y=pd.Series([3, 2, 1])),
reference=None,
),
),
Expand All @@ -35,7 +36,8 @@ def test_column_distribution_metric_success(
report = Report(metrics=[metric])
report.run(current_data=current_dataset, reference_data=reference_dataset, column_mapping=data_mapping)
result = metric.get_result()
assert result == expected_result
assert list(result.current.x) == list(expected_result.current.x)
assert list(result.current.y) == list(expected_result.current.y)


@pytest.mark.parametrize(
Expand Down Expand Up @@ -76,7 +78,7 @@ def test_column_distribution_metric_value_error(
pd.DataFrame({"col": [1, 2, 3]}),
None,
ColumnDistributionMetric(column_name="col"),
{"column_name": "col", "current": {"1": 1, "2": 1, "3": 1}, "reference": None},
{"column_name": "col"},
),
(
pd.DataFrame({"col1": [1, 2, 3], "col2": [10, 20, 3.5]}),
Expand All @@ -87,7 +89,7 @@ def test_column_distribution_metric_value_error(
}
),
ColumnDistributionMetric(column_name="col1"),
{"column_name": "col1", "current": {"1": 1, "2": 1, "3": 1}, "reference": {"10.0": 1, "20.0": 1, "3.5": 1}},
{"column_name": "col1"},
),
),
)
Expand Down