Follow us on LinkedIn  for project updates and announcements

Module: confusion_matrix_plot

Expand source code 
# Copyright (C) 2023-present The Project Contributors
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#    http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

from dataclasses import dataclass
from typing import List
from typing import Tuple
import numpy as np
import pandas as pd
from matplotlib import pyplot as plt
from matplotlib.colors import LinearSegmentedColormap
from cl.runtime import Context
from cl.runtime.plots.confusion_matrix_plot_style import ConfusionMatrixPlotStyle
from cl.runtime.plots.matplotlib_plot import MatplotlibPlot
from cl.runtime.plots.matplotlib_util import MatplotlibUtil
from cl.runtime.plots.matrix_util import MatrixUtil
from cl.runtime.records.dataclasses_extensions import field


@dataclass(slots=True, kw_only=True)
class ConfusionMatrixPlot(MatplotlibPlot):
    """Confusion matrix visualization for a categorical experiment."""

    title: str = field()
    """Plot title."""

    received_categories: List[str] = field()
    """List of received (predicted) categories for each trial."""

    expected_categories: List[str] = field()
    """List of expected (correct) categories in the same order of trials as received (predicted) categories."""

    x_label: str | None = "Predicted"
    """x-axis label."""

    y_label: str | None = "Correct"
    """y-axis label."""

    def _create_figure(self) -> plt.Figure:
        # Load style object or create with default settings if not specified
        style = self._load_style()
        theme = self._get_pyplot_theme(style=style)

        # TODO: consider moving
        data, annotation_text = self._create_confusion_matrix()

        with plt.style.context(theme):
            fig, axes = plt.subplots()

            cmap = LinearSegmentedColormap.from_list("rg", ["g", "y", "r"], N=256)

            im = MatplotlibUtil.heatmap(data.values, data.index.tolist(), data.columns.tolist(), ax=axes, cmap=cmap)
            MatplotlibUtil.annotate_heatmap(im, labels=annotation_text, textcolors="black", size=style.label_font_size)

            # Set figure and axes labels
            axes.set_xlabel(self.x_label)
            axes.set_ylabel(self.y_label)
            axes.set_title(self.title)

            fig.tight_layout()

        return fig

    def _load_style(self) -> ConfusionMatrixPlotStyle:
        """Load style object or create with default settings if not specified."""
        style = Context.current().load_one(ConfusionMatrixPlotStyle, self.style, is_key_optional=True)
        if style is None:
            # Use default values if not found
            style = ConfusionMatrixPlotStyle(plot_style_id="Default")
            style.init_all()
        return style

    def _create_confusion_matrix(self) -> Tuple[pd.DataFrame, List[List[str]]]:
        raw_data = pd.DataFrame({"Actual": self.expected_categories, "Predicted": self.received_categories})

        data_confusion_matrix = MatrixUtil.create_confusion_matrix(
            data=raw_data, true_column_name="Actual", predicted_column_name="Predicted"
        )
        data_confusion_matrix_percent = MatrixUtil.convert_confusion_matrix_to_percent(data=data_confusion_matrix)
        diag_mask = np.eye(data_confusion_matrix_percent.shape[0], dtype=bool)
        data_confusion_matrix_error_percent = data_confusion_matrix_percent.copy()
        data_confusion_matrix_error_percent.values[diag_mask] = 100 - np.diag(data_confusion_matrix_percent)
        annotation_text = MatrixUtil.create_confusion_matrix_labels(data=data_confusion_matrix, in_percent=True)

        return data_confusion_matrix_error_percent, annotation_text

Classes

class ConfusionMatrixPlot (*, plot_id: str = None, style: PlotStyleKey | None = None, title: str = None, received_categories: List[str] = None, expected_categories: List[str] = None, x_label: str | None = 'Predicted', y_label: str | None = 'Correct')

Confusion matrix visualization for a categorical experiment.

Expand source code 
@dataclass(slots=True, kw_only=True)
class ConfusionMatrixPlot(MatplotlibPlot):
    """Confusion matrix visualization for a categorical experiment."""

    title: str = field()
    """Plot title."""

    received_categories: List[str] = field()
    """List of received (predicted) categories for each trial."""

    expected_categories: List[str] = field()
    """List of expected (correct) categories in the same order of trials as received (predicted) categories."""

    x_label: str | None = "Predicted"
    """x-axis label."""

    y_label: str | None = "Correct"
    """y-axis label."""

    def _create_figure(self) -> plt.Figure:
        # Load style object or create with default settings if not specified
        style = self._load_style()
        theme = self._get_pyplot_theme(style=style)

        # TODO: consider moving
        data, annotation_text = self._create_confusion_matrix()

        with plt.style.context(theme):
            fig, axes = plt.subplots()

            cmap = LinearSegmentedColormap.from_list("rg", ["g", "y", "r"], N=256)

            im = MatplotlibUtil.heatmap(data.values, data.index.tolist(), data.columns.tolist(), ax=axes, cmap=cmap)
            MatplotlibUtil.annotate_heatmap(im, labels=annotation_text, textcolors="black", size=style.label_font_size)

            # Set figure and axes labels
            axes.set_xlabel(self.x_label)
            axes.set_ylabel(self.y_label)
            axes.set_title(self.title)

            fig.tight_layout()

        return fig

    def _load_style(self) -> ConfusionMatrixPlotStyle:
        """Load style object or create with default settings if not specified."""
        style = Context.current().load_one(ConfusionMatrixPlotStyle, self.style, is_key_optional=True)
        if style is None:
            # Use default values if not found
            style = ConfusionMatrixPlotStyle(plot_style_id="Default")
            style.init_all()
        return style

    def _create_confusion_matrix(self) -> Tuple[pd.DataFrame, List[List[str]]]:
        raw_data = pd.DataFrame({"Actual": self.expected_categories, "Predicted": self.received_categories})

        data_confusion_matrix = MatrixUtil.create_confusion_matrix(
            data=raw_data, true_column_name="Actual", predicted_column_name="Predicted"
        )
        data_confusion_matrix_percent = MatrixUtil.convert_confusion_matrix_to_percent(data=data_confusion_matrix)
        diag_mask = np.eye(data_confusion_matrix_percent.shape[0], dtype=bool)
        data_confusion_matrix_error_percent = data_confusion_matrix_percent.copy()
        data_confusion_matrix_error_percent.values[diag_mask] = 100 - np.diag(data_confusion_matrix_percent)
        annotation_text = MatrixUtil.create_confusion_matrix_labels(data=data_confusion_matrix, in_percent=True)

        return data_confusion_matrix_error_percent, annotation_text

Ancestors

Static methods

def get_key_type() -> Type

Inherited from: MatplotlibPlot.get_key_type

Return key type even when called from a record.

Fields

var expected_categories -> List[str]

List of expected (correct) categories in the same order of trials as received (predicted) categories.

var plot_id -> str

Inherited from: MatplotlibPlot.plot_id

Unique plot identifier.

var received_categories -> List[str]

List of received (predicted) categories for each trial.

var style -> PlotStyleKey | None

Inherited from: MatplotlibPlot.style

Color and layout options.

var title -> str

Plot title.

var x_label -> str | None

x-axis label.

var y_label -> str | None

y-axis label.

Methods

def get_key(self) -> PlotKey

Inherited from: MatplotlibPlot.get_key

Return a new key object whose fields populated from self, do not return self.

def get_view(self) -> View

Inherited from: MatplotlibPlot.get_view

Return a view object for the plot, implement using ‘create_figure’ method.

def init_all(self) -> None

Inherited from: MatplotlibPlot.init_all

Invoke ‘init’ for each class in the order from base to derived, then validate against schema.

def save_png(self) -> None

Inherited from: MatplotlibPlot.save_png

Save in png format to ‘base_dir/plot_id.png’, implement using ‘create_figure’ method.