diff --git a/figures.ipynb b/figures.ipynb
index 8187dba0f45838e964cfcd51d825f3d819038047..3fa61dd08eb55d438572eeae74a30378672978bd 100644
--- a/figures.ipynb
+++ b/figures.ipynb
@@ -26,7 +26,9 @@
"import pickle\n",
"\n",
"import holoviews.plotting.mpl\n",
- "import matplotlib; matplotlib.use(\"agg\")\n",
+ "import matplotlib\n",
+ "\n",
+ "matplotlib.use(\"agg\")\n",
"import matplotlib.pyplot as plt\n",
"import matplotlib.tri as mtri\n",
"import numpy as np\n",
@@ -61,7 +63,26 @@
"}\n",
"\n",
"plt.rcParams.update(params)\n",
- "plt.rc(\"text.latex\", preamble=[r\"\\usepackage{xfrac}\", r\"\\usepackage{siunitx}\"])"
+ "plt.rc(\"text.latex\", preamble=[r\"\\usepackage{xfrac}\", r\"\\usepackage{siunitx}\"])\n",
+ "\n",
+ "\n",
+ "class HistogramNormalize(matplotlib.colors.Normalize):\n",
+ " def __init__(self, data, vmin=None, vmax=None, mixing_degree=1):\n",
+ " self.mixing_degree = mixing_degree\n",
+ " if vmin is not None:\n",
+ " data = data[data >= vmin]\n",
+ " if vmax is not None:\n",
+ " data = data[data <= vmax]\n",
+ "\n",
+ " self.sorted_data = np.sort(data.flatten())\n",
+ " matplotlib.colors.Normalize.__init__(self, vmin, vmax)\n",
+ "\n",
+ " def __call__(self, value, clip=None):\n",
+ " hist_norm = np.ma.masked_array(\n",
+ " np.searchsorted(self.sorted_data, value) / len(self.sorted_data)\n",
+ " )\n",
+ " linear_norm = super().__call__(value, clip)\n",
+ " return self.mixing_degree * hist_norm + (1 - self.mixing_degree) * linear_norm"
]
},
{
@@ -315,16 +336,22 @@
" ax.triplot(triang, c=\"w\", lw=0.2, alpha=0.8)\n",
"\n",
" values = np.array(list(learner.data.values()))\n",
- " ax.imshow(\n",
- " learner.plot(npoints if kind == \"homogeneous\" else None).Image.I.data,\n",
+ " plot_data = learner.plot(npoints if kind == \"homogeneous\" else None).Image.I.data\n",
+ " im = ax.imshow(\n",
+ " plot_data,\n",
" extent=(-0.5, 0.5, -0.5, 0.5),\n",
" interpolation=\"none\",\n",
" )\n",
" ax.set_xticks([])\n",
" ax.set_yticks([])\n",
"\n",
+ " if i in [2, 3]:\n",
+ " norm = HistogramNormalize(plot_data, mixing_degree=0.6)\n",
+ " im.set_norm(norm)\n",
+ "\n",
"axs[0][0].set_ylabel(r\"$\\textrm{homogeneous}$\")\n",
"axs[1][0].set_ylabel(r\"$\\textrm{adaptive}$\")\n",
+ "\n",
"plt.savefig(\"figures/Learner2D.pdf\", bbox_inches=\"tight\", transparent=True)"
]
},
@@ -390,7 +417,7 @@
" axs.scatter(xs, ys, alpha=1, zorder=2, c=\"k\")\n",
" (x_left, x_right), loss = list(learner.losses.items())[\n",
" 0\n",
- " ] # it's a ItemSortedDict\n",
+ " ] # it's an ItemSortedDict\n",
" (y_left, y_right) = [\n",
" learner.data[x_left] + offset,\n",
" learner.data[x_right] + offset,\n",
@@ -540,7 +567,7 @@
"for i, ax in enumerate(axs):\n",
" ax.axis(\"off\")\n",
" ax.set_ylim(-1.5, 1.5)\n",
- " label = \"abcde\"[i]\n",
+ " label = \"abcdefgh\"[i]\n",
" ax.text(\n",
" 0.5,\n",
" 0.9,\n",
@@ -551,36 +578,35 @@
" )\n",
"\n",
"\n",
- "def plot_tri(xs, ax):\n",
+ "def plot_tri(xs, ax, colors):\n",
" ys = f(xs)\n",
" for i in range(len(xs)):\n",
" if i == 0 or i == len(xs) - 1:\n",
" continue\n",
- " color = f\"C{i}\"\n",
" verts = [(xs[i - 1], ys[i - 1]), (xs[i], ys[i]), (xs[i + 1], ys[i + 1])]\n",
- " poly = Polygon(verts, facecolor=color, alpha=0.4)\n",
+ " poly = Polygon(verts, facecolor=colors[xs[i]], alpha=0.4)\n",
" ax.add_patch(poly)\n",
- " ax.scatter([xs[i]], [ys[i]], c=color, s=6, zorder=11)\n",
+ " ax.scatter([xs[i]], [ys[i]], c=colors[xs[i]], s=6, zorder=11)\n",
" ax.plot(\n",
" [xs[i - 1], xs[i + 1]], [ys[i - 1], ys[i + 1]], c=\"k\", ls=\"--\", alpha=0.3\n",
" )\n",
"\n",
"\n",
- "plot_tri(xs, axs[1])\n",
+ "learner = adaptive.Learner1D(\n",
+ " f,\n",
+ " (xs[0], xs[-1]),\n",
+ " loss_per_interval=adaptive.learner.learner1D.curvature_loss_function(),\n",
+ ")\n",
+ "learner.tell_many(xs, ys)\n",
"\n",
- "for i in [2, 3, 4]:\n",
- " ax = axs[i]\n",
- " learner = adaptive.Learner1D(\n",
- " f,\n",
- " (xs[0], xs[-1]),\n",
- " loss_per_interval=adaptive.learner.learner1D.curvature_loss_function(),\n",
- " )\n",
- " learner.tell_many(xs, ys)\n",
- " x_new = learner.ask(1)[0][0]\n",
- " learner.tell(x_new, f(x_new))\n",
- " xs, ys = zip(*sorted(learner.data.items()))\n",
+ "for i, ax in enumerate(axs[1:]):\n",
+ " if i != 0:\n",
+ " x_new = learner.ask(1)[0][0]\n",
+ " learner.tell(x_new, f(x_new))\n",
+ " xs, ys = zip(*sorted(learner.data.items()))\n",
" plot(xs, ax)\n",
- " plot_tri(xs, ax)\n",
+ " colors = {x: f\"C{i}\" for i, x in enumerate(learner.data.keys())}\n",
+ " plot_tri(xs, ax, colors=colors)\n",
"\n",
"plt.savefig(\"figures/line_loss.pdf\", bbox_inches=\"tight\", transparent=True)\n",
"plt.show()"
@@ -716,13 +742,13 @@
" color = f\"C{j}\"\n",
" label = \"abc\"[j]\n",
" label = f\"$\\mathrm{{({label})}}$ {title}\"\n",
- "# ax.loglog(Ns, err_hom, ls=\"--\", c=color)\n",
- "# ax.loglog(Ns, err_adaptive, label=label, c=color)\n",
- " error = np.array(err_hom) / np.array(err_adaptive)\n",
- " if i == 0:\n",
- " ax.loglog(Ns[:36], error[:36], c=color, label=label)\n",
- " else:\n",
- " ax.loglog(Ns, error, c=color, label=label)\n",
+ " ax.loglog(Ns, err_hom, ls=\"--\", c=color)\n",
+ " ax.loglog(Ns, err_adaptive, label=label, c=color)\n",
+ "# error = np.array(err_hom) / np.array(err_adaptive)\n",
+ "# if i == 0:\n",
+ "# ax.loglog(Ns[:36], error[:36], c=color, label=label)\n",
+ "# else:\n",
+ "# ax.loglog(Ns, error, c=color, label=label)\n",
" ax.legend()\n",
"\n",
"plt.savefig(\"figures/line_loss_error.pdf\", bbox_inches=\"tight\", transparent=True)\n",