update code to work with more recent versions of dependencies

notebooks/quickstart_classification_iris.py

@@ -152,10 +152,10 @@
 # This is to indicate which quantity should be used for model selection.
 
 # %%
-from bokeh.plotting import figure, Figure
+from bokeh.plotting import figure
 
 
-def make_figure(title: str) -> Figure:
+def make_figure(title: str) -> figure:
     p = figure(title=title)
     p.line([0, 1, 2, 3], np.random.random(size=4), line_width=2)
     return p

src/metriculous/comparison_test.py

@@ -4,12 +4,12 @@
 
 import pytest
 from bokeh import plotting
-from bokeh.plotting import Figure
+from bokeh.plotting import figure
 
 from metriculous import Comparison, Evaluation, Quantity
 
 
-def make_a_bokeh_figure() -> Figure:
+def make_a_bokeh_figure() -> figure:
     p = plotting.figure()
     p.line(x=[0, 1, 5, 6], y=[40, 60, 30, 50])
     return p

src/metriculous/evaluation.py

@@ -6,7 +6,7 @@
 from dataclasses import dataclass, field, replace
 from typing import Callable, Generic, Optional, Sequence, TypeVar, Union
 
-from bokeh.plotting import Figure
+from bokeh.plotting import figure
 
 
 @dataclass(frozen=True)
@@ -21,7 +21,7 @@ class Quantity:
 class Evaluation:
     model_name: str
     quantities: Sequence[Quantity] = field(default_factory=list)
-    lazy_figures: Sequence[Callable[[], Figure]] = field(default_factory=list)
+    lazy_figures: Sequence[Callable[[], figure]] = field(default_factory=list)
     primary_metric: Optional[str] = None
 
     def get_by_name(self, quantity_name: str) -> Quantity:
@@ -37,7 +37,7 @@ def get_primary(self) -> Optional[Quantity]:
             return None
         return self.get_by_name(self.primary_metric)
 
-    def figures(self) -> Sequence[Figure]:
+    def figures(self) -> Sequence[figure]:
         return [f() for f in self.lazy_figures]
 
     def filtered(

src/metriculous/evaluators/bokeh_utils.py

@@ -3,7 +3,7 @@
 import numpy as np
 from bokeh.embed import file_html
 from bokeh.models import Div, Title
-from bokeh.plotting import Figure
+from bokeh.plotting import figure
 from bokeh.resources import CDN
 
 TOOLS = "pan,box_zoom,wheel_zoom,reset"
@@ -37,15 +37,15 @@ def title_div(title_rows: Sequence[str]) -> Div:
     )
 
 
-def add_title_rows(p: Figure, title_rows: Sequence[str]) -> None:
+def add_title_rows(p: figure, title_rows: Sequence[str]) -> None:
     for title_row in reversed(title_rows):
         p.add_layout(
             Title(text=title_row, text_font_size=FONT_SIZE, align="center"),
             place="above",
         )
 
 
-def apply_default_style(p: Figure) -> None:
+def apply_default_style(p: figure) -> None:
     p.background_fill_color = BACKGROUND_COLOR
     p.grid.grid_line_color = "white"
 
@@ -80,7 +80,7 @@ def scatter_plot_circle_size(
     return max(smallest, biggest - slope * num_points)
 
 
-def check_that_all_figures_can_be_rendered(figures: Sequence[Figure]) -> None:
+def check_that_all_figures_can_be_rendered(figures: Sequence[figure]) -> None:
     """Generates HTML for each figure.
 
     In some cases this reveals issues that might not be noticed if we just instantiated the figures

src/metriculous/evaluators/classification/calibration_plot.py

@@ -4,7 +4,7 @@
 import numpy as np
 from bokeh import plotting
 from bokeh.models import ColumnDataSource, HoverTool
-from bokeh.plotting import Figure
+from bokeh.plotting import figure
 from sklearn.utils import check_consistent_length, column_or_1d
 
 from metriculous.evaluators.bokeh_utils import (
@@ -23,7 +23,7 @@ def _bokeh_probability_calibration_plot(
     y_pred_score: np.ndarray,
     title_rows: Sequence[str],
     # TODO sample_weights: Optional[np.ndarray],
-) -> Callable[[], Figure]:
+) -> Callable[[], figure]:
     """Probability calibration plot.
 
     Args:
@@ -39,7 +39,7 @@ def _bokeh_probability_calibration_plot(
 
     """
 
-    def figure() -> Figure:
+    def create_figure() -> figure:
         assert y_true_binary.shape == y_pred_score.shape
         assert set(y_true_binary).issubset({0, 1}) or set(y_true_binary).issubset(
             {False, True}
@@ -71,8 +71,8 @@ def figure() -> Figure:
         )
 
         p = plotting.figure(
-            # plot_height=370,
-            # plot_width=350,
+            # height=370,
+            # width=350,
             x_range=(-0.05, 1.05),
             y_range=(-0.05, 1.05),
             tools=TOOLS,
@@ -121,7 +121,7 @@ def figure() -> Figure:
 
         return p
 
-    return figure
+    return create_figure
 
 
 @dataclass(frozen=True)

src/metriculous/evaluators/classification/classification_evaluator.py

@@ -3,7 +3,7 @@
 import numpy as np
 import numpy.testing as npt
 from assertpy import assert_that
-from bokeh.plotting import Figure
+from bokeh.plotting import figure
 from scipy.stats import entropy
 from sklearn import metrics as sklmetrics
 
@@ -203,9 +203,9 @@ def _lazy_figures(
         data: ClassificationData,
         maybe_sample_weights: Optional[np.ndarray],
         class_names: Sequence[str],
-    ) -> Sequence[Tuple[str, Callable[[], Figure]]]:
+    ) -> Sequence[Tuple[str, Callable[[], figure]]]:
 
-        lazy_figures: List[Tuple[str, Callable[[], Figure]]] = []
+        lazy_figures: List[Tuple[str, Callable[[], figure]]] = []
 
         y_true = data.target.argmaxes
         y_true_one_hot = data.target.argmaxes_one_hot

src/metriculous/evaluators/classification/classification_evaluator_test.py

@@ -219,7 +219,7 @@ def test_ClassificationEvaluator_perfect_prediction(
         ),
         Quantity(
             name="Log Loss",
-            value=2.1094237467877998e-15,
+            value=2.220446e-16,
             higher_is_better=False,
             description=None,
         ),

src/metriculous/evaluators/classification/classification_figures_bokeh.py

@@ -4,7 +4,7 @@
 from assertpy import assert_that
 from bokeh import plotting
 from bokeh.models import ColumnDataSource, HoverTool, LinearColorMapper
-from bokeh.plotting import Figure
+from bokeh.plotting import figure
 from sklearn import metrics as sklmetrics
 
 from metriculous.evaluators.bokeh_utils import (
@@ -33,7 +33,7 @@ def _bokeh_output_histogram(
     title_rows: Sequence[str],
     sample_weights: Optional[np.ndarray] = None,
     x_label_rotation: Union[str, float] = "horizontal",
-) -> Callable[[], Figure]:
+) -> Callable[[], figure]:
     """Histogram of ground truth and prediction.
 
     Args:
@@ -60,15 +60,15 @@ def _bokeh_output_histogram(
     assert_that(np.shape(y_true)).is_equal_to(np.shape(weights))
     normalize = not np.allclose(weights, 1.0)
 
-    def figure() -> Figure:
+    def create_figure() -> figure:
         n = len(class_names)
 
         bins = np.arange(0, n + 1, 1)
 
         p = plotting.figure(
             x_range=class_names,
-            plot_height=350,
-            plot_width=350,
+            height=350,
+            width=350,
             tools=TOOLS,
             toolbar_location=TOOLBAR_LOCATION,
         )
@@ -110,7 +110,7 @@ def figure() -> Figure:
         p.x_range.bounds = (-0.5, 0.5 + len(class_names))
         return p
 
-    return figure
+    return create_figure
 
 
 def _bokeh_confusion_matrix(
@@ -120,7 +120,7 @@ def _bokeh_confusion_matrix(
     title_rows: Sequence[str],
     x_label_rotation: Union[str, float] = "horizontal",
     y_label_rotation: Union[str, float] = "vertical",
-) -> Callable[[], Figure]:
+) -> Callable[[], figure]:
     """Confusion matrix heatmap.
 
     Args:
@@ -142,7 +142,7 @@ def _bokeh_confusion_matrix(
 
     """
 
-    def figure() -> Figure:
+    def create_figure() -> figure:
         cm = sklmetrics.confusion_matrix(
             y_true, y_pred, labels=list(range(len(class_names)))
         )
@@ -241,7 +241,7 @@ def figure() -> Figure:
 
         return p
 
-    return figure
+    return create_figure
 
 
 def _bokeh_confusion_scatter(
@@ -251,7 +251,7 @@ def _bokeh_confusion_scatter(
     title_rows: Sequence[str],
     x_label_rotation: Union[str, float] = "horizontal",
     y_label_rotation: Union[str, float] = "vertical",
-) -> Callable[[], Figure]:
+) -> Callable[[], figure]:
     """Scatter plot that contains the same information as a confusion matrix.
 
     Args:
@@ -273,15 +273,15 @@ def _bokeh_confusion_scatter(
 
     """
 
-    def figure() -> Figure:
+    def create_figure() -> figure:
         if len(y_true) != len(y_pred):
             raise ValueError("y_true and y_pred must have the same length!")
 
         p = plotting.figure(
             x_range=(-0.5, -0.5 + len(class_names)),
             y_range=(-0.5, -0.5 + len(class_names)),
-            plot_height=350,
-            plot_width=350,
+            height=350,
+            width=350,
             tools=TOOLS,
             toolbar_location=TOOLBAR_LOCATION,
             match_aspect=True,
@@ -334,15 +334,15 @@ def noise() -> np.ndarray:
 
         return p
 
-    return figure
+    return create_figure
 
 
 def _bokeh_roc_curve(
     y_true_binary: np.ndarray,
     y_pred_score: np.ndarray,
     title_rows: Sequence[str],
     sample_weights: Optional[np.ndarray],
-) -> Callable[[], Figure]:
+) -> Callable[[], figure]:
     """Interactive receiver operator characteristic (ROC) curve.
 
     Args:
@@ -360,7 +360,7 @@ def _bokeh_roc_curve(
 
     """
 
-    def figure() -> Figure:
+    def create_figure() -> figure:
         assert y_true_binary.shape == y_pred_score.shape
         assert set(y_true_binary).issubset({0, 1}) or set(y_true_binary).issubset(
             {False, True}
@@ -381,8 +381,8 @@ def figure() -> Figure:
         )
 
         p = plotting.figure(
-            plot_height=400,
-            plot_width=350,
+            height=400,
+            width=350,
             tools=TOOLS,
             toolbar_location=TOOLBAR_LOCATION,
             # toolbar_location=None,  # hides entire toolbar
@@ -422,15 +422,15 @@ def figure() -> Figure:
 
         return p
 
-    return figure
+    return create_figure
 
 
 def _bokeh_precision_recall_curve(
     y_true_binary: np.ndarray,
     y_pred_score: np.ndarray,
     title_rows: Sequence[str],
     sample_weights: Optional[np.ndarray],
-) -> Callable[[], Figure]:
+) -> Callable[[], figure]:
     """Interactive precision recall curve.
 
     Args:
@@ -448,7 +448,7 @@ def _bokeh_precision_recall_curve(
 
     """
 
-    def figure() -> Figure:
+    def create_figure() -> figure:
         assert y_true_binary.shape == y_pred_score.shape
         assert set(y_true_binary).issubset({0, 1}) or set(y_true_binary).issubset(
             {False, True}
@@ -464,8 +464,8 @@ def figure() -> Figure:
         recall = recall[:-1]
 
         p = plotting.figure(
-            plot_height=400,
-            plot_width=350,
+            height=400,
+            width=350,
             x_range=(-0.05, 1.05),
             y_range=(-0.05, 1.05),
             tools=TOOLS,
@@ -499,15 +499,15 @@ def figure() -> Figure:
 
         return p
 
-    return figure
+    return create_figure
 
 
 def _bokeh_automation_rate_analysis(
     y_target_one_hot: np.ndarray,
     y_pred_proba: np.ndarray,
     title_rows: Sequence[str],
     sample_weights: Optional[np.ndarray],
-) -> Callable[[], Figure]:
+) -> Callable[[], figure]:
     """
     Plots various quantities over automation rate, where a single probability threshold
     is used for all classes to decide if we are confident enough to automate the
@@ -528,7 +528,7 @@ def _bokeh_automation_rate_analysis(
 
     """
 
-    def figure() -> Figure:
+    def create_figure() -> figure:
         # ----- Check input -----
         assert y_target_one_hot.ndim == 2
         assert y_pred_proba.ndim == 2
@@ -572,8 +572,8 @@ def figure() -> Figure:
 
         # ----- Bokeh plot -----
         p = plotting.figure(
-            plot_height=400,
-            plot_width=350,
+            height=400,
+            width=350,
             x_range=(-0.05, 1.05),
             y_range=(-0.05, 1.05),
             tools=TOOLS,
@@ -634,7 +634,7 @@ def figure() -> Figure:
 
         return p
 
-    return figure
+    return create_figure
 
 
 def _faster_accuracy(