Spaces:
Running
Running
File size: 10,384 Bytes
11ac28c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 |
{
"cells": [
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"from pathlib import Path\n",
"\n",
"import pandas as pd\n",
"import plotly.colors as pcolors\n",
"import seaborn as sns\n",
"from matplotlib import pyplot as plt\n",
"\n",
"from mlip_arena.models import MLIPEnum\n",
"\n",
"mlip_methods = [\n",
" model.name\n",
" for model in MLIPEnum\n",
"]\n",
"\n",
"all_attributes = dir(pcolors.qualitative)\n",
"color_palettes = {\n",
" attr: getattr(pcolors.qualitative, attr)\n",
" for attr in all_attributes\n",
" if isinstance(getattr(pcolors.qualitative, attr), list)\n",
"}\n",
"color_palettes.pop(\"__all__\", None)\n",
"\n",
"palette_names = list(color_palettes.keys())\n",
"palette_colors = list(color_palettes.values())\n",
"palette_name = \"Plotly\"\n",
"color_sequence = color_palettes[palette_name] # type: ignore\n",
"\n",
"method_color_mapping = {\n",
" method: color_sequence[i % len(color_sequence)]\n",
" for i, method in enumerate(mlip_methods)\n",
"}"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": [
"%matplotlib inline\n",
"\n",
"import numpy as np\n",
"\n",
"from mlip_arena.models import MLIPEnum\n",
"\n",
"# Color mapping by class\n",
"color_mapping = {\n",
" \"DAC\": \"#e41a1c\",\n",
" \"Flue Gas\": \"#377eb8\",\n",
" \"General\": \"#4daf4a\"\n",
"}\n",
"\n",
"# Decision boundary thresholds\n",
"thresholds = {\n",
" \"General\": (None, 35),\n",
" \"Flue Gas\": (35, 50),\n",
" \"DAC\": (50, 100)\n",
"}\n",
"\n",
"# Collect data from all models\n",
"all_data = []\n",
"margins = []\n",
"\n",
"for model in MLIPEnum:\n",
" fpath = Path(f\"{model.name}.pkl\")\n",
" if not fpath.exists():\n",
" continue\n",
"\n",
" df = pd.read_pickle(fpath)\n",
" df = df.drop_duplicates(subset=[\"model\", \"name\", \"class\"], keep=\"last\")\n",
" df_exploded = df.explode([\"henry_coefficient\", \"averaged_interaction_energy\", \"heat_of_adsorption\"])\n",
" df_group = df_exploded.groupby([\"model\", \"name\", \"class\"])[[\"henry_coefficient\", \"averaged_interaction_energy\", \"heat_of_adsorption\"]].mean().reset_index()\n",
"\n",
" df_group[\"model_name\"] = model.name\n",
" df_group[\"neg_heat\"] = -df_group[\"heat_of_adsorption\"] # negate for log scale\n",
" df_group = df_group[df_group[\"neg_heat\"] > 0] # remove invalid values\n",
"\n",
" df_group = df_group[df_group[\"name\"] != \"MIL-96-Al\"]\n",
"\n",
" all_data.append(df_group)\n",
"\n",
" # Compute misclassification margin\n",
" def point_misclassified(row):\n",
" val = row[\"neg_heat\"]\n",
" lower, upper = thresholds[row[\"class\"]]\n",
" return (lower is not None and val < lower) or (upper is not None and val >= upper)\n",
"\n",
" misclassified = df_group[df_group.apply(point_misclassified, axis=1)]\n",
"\n",
" def distance_to_boundary(row):\n",
" val = row[\"neg_heat\"]\n",
" lower, upper = thresholds[row[\"class\"]]\n",
" distances = []\n",
" if lower is not None:\n",
" distances.append(abs(val - lower))\n",
" if upper is not None:\n",
" distances.append(abs(val - upper))\n",
" return min(distances)\n",
"\n",
" if not misclassified.empty:\n",
" num_misclassified = len(misclassified) + (18 - len(df_group))\n",
" margin = misclassified.apply(distance_to_boundary, axis=1).mean()\n",
" else:\n",
" num_misclassified = 0\n",
" margin = 0.0\n",
"\n",
" margins.append((model.name, margin, num_misclassified))\n",
"\n",
"\n",
"# Combine all into one DataFrame\n",
"combined_df = pd.concat(all_data, ignore_index=True)\n",
"margins_df = pd.DataFrame(margins, columns=[\"model_name\", \"misclassification_margin\", \"num_misclassified\"])\n",
"\n",
"# --- Plotting ---\n",
"\n",
"with plt.style.context(\"default\"):\n",
"\n",
" LARGE_SIZE = 10\n",
" MEDIUM_SIZE = 8\n",
" SMALL_SIZE = 6\n",
"\n",
" plt.rcParams.update({\n",
" \"font.size\": SMALL_SIZE,\n",
" \"axes.titlesize\": MEDIUM_SIZE,\n",
" \"axes.labelsize\": MEDIUM_SIZE,\n",
" \"xtick.labelsize\": SMALL_SIZE,\n",
" \"ytick.labelsize\": SMALL_SIZE,\n",
" \"legend.fontsize\": SMALL_SIZE,\n",
" \"figure.titlesize\": LARGE_SIZE,\n",
" })\n",
"\n",
" fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(6, 4), sharex=False, gridspec_kw={\"height_ratios\": [3, 1.5]})\n",
"\n",
" # --- Main Stripplot ---\n",
" sns.stripplot(\n",
" data=combined_df,\n",
" x=\"neg_heat\",\n",
" y=\"model_name\",\n",
" hue=\"class\",\n",
" size=2,\n",
" palette=color_mapping,\n",
" dodge=True,\n",
" jitter=0.1,\n",
" alpha=1,\n",
" ax=ax1,\n",
" )\n",
"\n",
" xmin, xmax = ax1.get_xlim()\n",
"\n",
" ax1.axvspan(xmin, 35, color=color_mapping[\"General\"], alpha=0.1, label=\"General\")\n",
" ax1.axvspan(35, 50, color=color_mapping[\"Flue Gas\"], alpha=0.1, label=\"Flue Gas\")\n",
" ax1.axvspan(50, 100, color=color_mapping[\"DAC\"], alpha=0.1, label=\"DAC\")\n",
"\n",
" ax1.axvline(x=35, linestyle=\"--\", color=\"gray\", label=\"Exp. $\\\\mathregular{CO_2}$ $Q_\\\\text{st}$ = 35 kJ/mol\")\n",
" ax1.axvline(x=50, linestyle=\"--\", color=\"gray\", label=\"Exp. $\\\\mathregular{CO_2}$ $Q_\\\\text{st}$ = 50 kJ/mol\")\n",
" ax1.axvline(x=100, linestyle=\"--\", color=\"gray\", label=\"Exp. $\\\\mathregular{CO_2}$ $Q_\\\\text{st}$ = 100 kJ/mol\")\n",
"\n",
" ax1.set_xscale(\"log\")\n",
" ax1.set_xlabel(\"Heat of $\\\\mathregular{CO_2}$ Adsorption $Q_\\\\text{st}$ [kJ/mol]\")\n",
" ax1.set_ylabel(\"\")\n",
" ax1.set_xlim(xmin, xmax)\n",
"\n",
" yticks = ax1.get_yticks()\n",
" yticks = np.array(yticks)\n",
" yticks = yticks[np.isfinite(yticks)] # Remove any NaNs\n",
"\n",
" # Draw horizontal lines between models (skip the last one)\n",
" for y in yticks[:-1] + np.diff(yticks) / 2:\n",
" ax1.axhline(y=y, color=\"gray\", linestyle=\":\", linewidth=0.7, alpha=0.5, zorder=0)\n",
"\n",
" handles, labels = ax1.get_legend_handles_labels()\n",
" legend_dict = dict(zip(labels, handles, strict=False))\n",
"\n",
" desired_order = [\n",
" \"General\", \"Exp. $\\\\mathregular{CO_2}$ $Q_\\\\text{st}$ = 35 kJ/mol\", \"Flue Gas\",\n",
" \"Exp. $\\\\mathregular{CO_2}$ $Q_\\\\text{st}$ = 50 kJ/mol\", \"DAC\", \"Exp. $\\\\mathregular{CO_2}$ $Q_\\\\text{st}$ = 100 kJ/mol\"\n",
" ]\n",
"\n",
" ordered_handles = [legend_dict[label] for label in desired_order if label in legend_dict]\n",
"\n",
" ax1.legend(\n",
" ordered_handles,\n",
" desired_order,\n",
" loc=\"lower center\",\n",
" bbox_to_anchor=(0.5, 1),\n",
" ncol=3,\n",
" frameon=True\n",
" )\n",
"\n",
"\n",
" ax1.spines[\"top\"].set_visible(False)\n",
" ax1.spines[\"right\"].set_visible(False)\n",
"\n",
" # --- Misclassification Margin Barplot ---\n",
"\n",
" # Sort by error margin\n",
" margins_df_sorted = margins_df.sort_values(by=\"misclassification_margin\", ascending=True)\n",
"\n",
" # Extract color values in order\n",
" bar_colors = [method_color_mapping[m] for m in margins_df_sorted[\"model_name\"]]\n",
"\n",
" sns.scatterplot(\n",
" data=margins_df_sorted,\n",
" x=\"num_misclassified\",\n",
" y=\"misclassification_margin\",\n",
" hue=\"model_name\",\n",
" palette=bar_colors,\n",
" ax=ax2\n",
" )\n",
"\n",
" for _, row in margins_df_sorted.iterrows():\n",
" x = row[\"num_misclassified\"]\n",
" y = row[\"misclassification_margin\"]\n",
" model = row[\"model_name\"]\n",
" color = bar_colors[margins_df_sorted[\"model_name\"].tolist().index(model)]\n",
"\n",
" ax2.text(\n",
" x+0.1,\n",
" y,\n",
" f\"{y:.2f}\",\n",
" fontsize=SMALL_SIZE,\n",
" ha=\"left\",\n",
" va=\"bottom\",\n",
" color=color,\n",
" alpha=0.9\n",
" )\n",
"\n",
" ax2.set_ylabel(\"Misclass. margin [kJ/mol]\")\n",
" ax2.set_xlabel(\"Missing + misclass. count\")\n",
" ax2.spines[\"top\"].set_visible(False)\n",
" ax2.spines[\"right\"].set_visible(False)\n",
" # ax2.set_xticklabels(margins_df_sorted[\"model_name\"], rotation=45)\n",
" ax2.set_yscale(\"log\")\n",
"\n",
" handles, labels = ax2.get_legend_handles_labels()\n",
" legend_dict = dict(zip(labels, handles, strict=False))\n",
" ax2.legend(\n",
" legend_dict.values(),\n",
" legend_dict.keys(),\n",
" loc=\"upper left\",\n",
" bbox_to_anchor=(0, 1),\n",
" ncol=3,\n",
" frameon=True\n",
" )\n",
"\n",
" plt.tight_layout()\n",
" plt.savefig(\"mof-misclassification_margin.pdf\", bbox_inches=\"tight\")\n",
" plt.show()\n"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "mlip-arena",
"language": "python",
"name": "mlip-arena"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.11.8"
}
},
"nbformat": 4,
"nbformat_minor": 2
}
|