Add task to create other illustrations

src/tranquilo_dev/plotting/illustrations.py

@@ -1,14 +1,16 @@
 from copy import deepcopy
 
+import estimagic as em
 import numpy as np
 import plotly.express as px
 import plotly.graph_objects as go
 from plotly.subplots import make_subplots
 from tranquilo.visualize import _clean_legend_duplicates
+from tranquilo.visualize import _get_sample_points
 
 
 def create_noise_plots():
-    out = {}
+    out = []
 
     def func(x):
         return x**4 + x**2 - x
@@ -57,7 +59,7 @@ def model(x, beta):
             "orientation": "h",
         }
     )
-    out[0] = fig  # fig.write_image("noise_plot_0.svg")
+    out.append(fig)  # fig.write_image("noise_plot_0.svg")
 
     fig1 = deepcopy(fig)
 
@@ -71,10 +73,10 @@ def model(x, beta):
             showlegend=False,
         )
     )
-    out[1] = fig1  # fig1.write_image("noise_plot_1.svg")
+    out.append(fig1)  # fig1.write_image("noise_plot_1.svg")
 
     plotting_data = []
-    for i, trustregion in enumerate([(-1.75, -0.5), (-0.75, 0.5)]):
+    for _, trustregion in enumerate([(-1.75, -0.5), (-0.75, 0.5)]):
         in_region = (trustregion[0] <= noise_x_grid) & (noise_x_grid <= trustregion[1])
         xs = noise_x_grid.copy()
         ys = noise_y.copy()
@@ -126,7 +128,7 @@ def model(x, beta):
         fig2.update_layout(width=600)
 
         plotting_data.append(fig2.data)
-        out[2 + i] = fig2  # fig2.write_image(f'noise_plot_{i+2}.svg')
+        out.append(fig2)  # fig2.write_image(f'noise_plot_{i+2}.svg')
 
     len(plotting_data)
     fig3 = make_subplots(cols=2, rows=1)
@@ -151,5 +153,236 @@ def model(x, beta):
         linecolor="black",
         zeroline=False,
     )
-    out[4] = fig3
+    out.append(fig3)
     return out
+
+
+def create_other_illustration_plots():
+    out = {}
+
+    def sphere(x):
+        return {"root_contributions": x}
+
+    res = em.minimize(
+        sphere,
+        params=np.array([1, 1]),
+        algorithm="tranquilo_ls",
+    )
+    params_history = np.array(res.history["params"])
+    states = res.algorithm_output["states"]
+    color_dict = {
+        "existing": "rgb(0,0,255)",
+        "new": "rgb(230,0,0)",
+        "discarded": "rgb(148, 148, 148)",
+    }
+
+    for k in range(6):
+        out[f"animation_{k}.svg"] = _plot_sample_points(
+            params_history, states, k, color_dict
+        )
+
+    base_fig = _plot_sample_points(params_history, states, 1, color_dict)
+    fig = deepcopy(base_fig)
+    center = np.array([fig.data[0].x[0], fig.data[0].y[0]])
+    candidate = np.array([fig.data[-1].x[0], fig.data[-1].y[0]])
+    direction = candidate - center
+    line_points = []
+    for i in [2, 4, 8]:
+        line_points.append(center + i * direction)
+    line_points = np.array(line_points)
+    fig.data[1].marker.color = fig.data[0].marker.color
+    fig.data[1].showlegend = False
+    fig.add_trace(
+        go.Scatter(
+            x=line_points[:, 0],
+            y=line_points[:, 1],
+            mode="markers",
+            marker_color="#805B87",
+            marker_size=9,
+            name="line search",
+        )
+    )
+    fig.update_xaxes(range=[0, 1.2])
+    fig.update_yaxes(range=[0, 1.2])
+    out["line_points_3.svg"] = deepcopy(fig)  # fig.write_image("line_points_3.svg")
+
+    fig.data[3].x = fig.data[3].x[:-1]
+    fig.data[3].y = fig.data[3].y[:-1]
+    out["line_points_2.svg"] = deepcopy(fig)  # fig.write_image("line_points_2.svg")
+
+    fig.data[3].x = fig.data[3].x[:-1]
+    fig.data[3].y = fig.data[3].y[:-1]
+    out["line_points_1.svg"] = deepcopy(fig)  # fig.write_image("line_points_1.svg")
+
+    fig.data = fig.data[:-1]
+    out["origin_plot.svg"] = deepcopy(fig)  # fig.write_image("origin_plot.svg")
+
+    center = candidate
+    radius = states[1].trustregion.radius
+    fig.add_shape(
+        type="circle",
+        xref="x",
+        yref="y",
+        x0=center[0] - radius,
+        y0=center[1] - radius,
+        x1=center[0] + radius,
+        y1=center[1] + radius,
+        line_width=0.5,
+        line_color="black",
+        line_dash="dash",
+    )
+    out["empty_speculative_trustregion_small_scale.svg"] = deepcopy(fig)
+
+    fig.update_yaxes(range=[0.5, 1.2])
+    fig.update_xaxes(range=[0.5, 1.2])
+    out["empty_speculative_trustregion_large_scale.svg"] = deepcopy(fig)
+
+    fig.update_yaxes(range=[0, 1.2])
+    fig.update_xaxes(range=[0, 1.2])
+
+    angle1 = 30
+    angle2 = 55
+    speculative1 = np.zeros((2, 2))
+    speculative1[0, 0] = candidate[0] - np.cos(angle1 * np.pi / 180) * radius
+    speculative1[0, 1] = candidate[1] - np.sin(angle1 * np.pi / 180) * radius
+    speculative1[1, 0] = candidate[0] + np.sin(angle2 * np.pi / 180) * radius
+    speculative1[1, 1] = candidate[1] - np.cos(angle2 * np.pi / 180) * radius
+    fig.data = fig.data[:3]
+    fig.add_trace(
+        go.Scatter(
+            x=speculative1[:, 0],
+            y=speculative1[:, 1],
+            mode="markers",
+            marker_color="#F98900",
+            marker_size=9,
+            name="speculative",
+        )
+    )
+    out["sampled_speculative_trustregion_small_scale.svg"] = deepcopy(fig)
+
+    fig.update_yaxes(range=[0.5, 1.2])
+    fig.update_xaxes(range=[0.5, 1.2])
+    out["sampled_speculative_trustregion_large_scale.svg"] = deepcopy(fig)
+
+    fig.update_yaxes(range=[0, 1.2])
+    fig.update_xaxes(range=[0, 1.2])
+    angle1 = 15
+    angle2 = 78
+    speculative1[0, 0] = candidate[0] + np.sin(angle1 * np.pi / 180) * radius
+    speculative1[0, 1] = candidate[1] - np.cos(angle1 * np.pi / 180) * radius
+    speculative1[1, 0] = candidate[0] + np.sin(angle2 * np.pi / 180) * radius
+    speculative1[1, 1] = candidate[1] - np.cos(angle2 * np.pi / 180) * radius
+
+    fig.data = fig.data[:3]
+    fig.add_trace(
+        go.Scatter(
+            x=speculative1[:, 0],
+            y=speculative1[:, 1],
+            mode="markers",
+            marker_color="#F98900",
+            marker_size=9,
+            name="speculative",
+        )
+    )
+    fig.add_trace(
+        go.Scatter(
+            x=line_points[:, 0],
+            y=line_points[:, 1],
+            mode="markers",
+            marker_color="#805B87",
+            marker_size=9,
+            name="line search",
+        )
+    )
+
+    out["line_and_speculative_points.svg"] = deepcopy(fig)
+    fig.update_yaxes(scaleanchor="x", scaleratio=1)
+    fig.update_xaxes(scaleanchor="y", scaleratio=1)
+
+    fig.update_layout(height=700, width=800, template="plotly_white")
+    out["checklayout.svg"] = deepcopy(fig)
+
+    return out
+
+
+# ======================================================================================
+# Shared functions
+# ======================================================================================
+
+
+def _plot_sample_points(history, states, iteration, color_dict):
+
+    state = states[iteration]
+    sample_points = _get_sample_points(state, history)
+    trustregion = state.trustregion
+    radius = trustregion.radius
+    center = trustregion.center
+    fig = go.Figure()
+    fig.add_shape(
+        type="circle",
+        xref="x",
+        yref="y",
+        x0=center[0] - radius,
+        y0=center[1] - radius,
+        x1=center[0] + radius,
+        y1=center[1] + radius,
+        line_width=0.5,
+        line_color="grey",
+    )
+
+    fig.add_traces(
+        px.scatter(
+            sample_points,
+            x=0,
+            y=1,
+            color="case",
+            color_discrete_map=color_dict,
+            opacity=0.7,
+        ).data,
+    )
+    if iteration >= 1:
+        fig.add_trace(
+            go.Scatter(
+                x=[state.candidate_x[0]],
+                y=[state.candidate_x[1]],
+                mode="markers",
+                marker_symbol="star",
+                marker_color="#228b22",
+                name="candidate",
+            ),
+        )
+    fig.update_traces(
+        marker_size=9,
+    )
+
+    fig = _clean_legend_duplicates(fig)
+    fig.update_layout(height=700, width=800, template="plotly_white")
+    fig.update_yaxes(
+        showgrid=False,
+        showline=True,
+        linewidth=1,
+        linecolor="black",
+        zeroline=False,
+        range=[-1.6, 1.6],
+    )
+    fig.update_xaxes(
+        showgrid=False,
+        showline=True,
+        linewidth=1,
+        linecolor="black",
+        zeroline=False,
+        range=[-1.6, 1.6],
+    )
+    fig.update_yaxes(scaleanchor="x", scaleratio=1)
+    fig.update_xaxes(scaleanchor="y", scaleratio=1)
+
+    layout = go.Layout(
+        margin=go.layout.Margin(
+            l=10,  # left margin
+            r=10,  # right margin
+            b=10,  # bottom margin
+            t=10,  # top margin
+        )
+    )
+    fig = fig.update_layout(layout)
+    return fig

src/tranquilo_dev/plotting/task_create_presentation_illustrations.py

@@ -0,0 +1,33 @@
+import pytask
+from tranquilo_dev.config import BLD
+from tranquilo_dev.plotting.illustrations import create_noise_plots
+from tranquilo_dev.plotting.illustrations import create_other_illustration_plots
+
+BLD_SLIDEV = BLD.joinpath("bld_slidev")
+
+
+@pytask.mark.produces({k: BLD_SLIDEV.joinpath(f"noise_plot_{k}.svg") for k in range(5)})
+def task_create_noise_plots(produces):
+    figures = create_noise_plots()
+    for path, fig in zip(produces.values(), figures):
+        fig.write_image(path)
+
+
+OTHER_ILLUSTRATION_NAMES = [
+    *[f"animation_{k}.svg" for k in range(6)],
+    *[f"line_points_{k}.svg" for k in range(1, 4)],
+    "origin_plot.svg",
+    "empty_speculative_trustregion_small_scale.svg",
+    "empty_speculative_trustregion_large_scale.svg",
+    "sampled_speculative_trustregion_small_scale.svg",
+    "sampled_speculative_trustregion_large_scale.svg",
+    "line_and_speculative_points.svg",
+    "checklayout.svg",
+]
+
+
+@pytask.mark.produces([BLD_SLIDEV.joinpath(name) for name in OTHER_ILLUSTRATION_NAMES])
+def task_create_other_illustration_plots(produces):
+    figures = create_other_illustration_plots()
+    for path in produces.values():
+        figures[path.name].write_image(path)