Merge pull request #87 from WMD-group/publication

Publication updates

Publication/README.md

@@ -0,0 +1,7 @@
+# Publications
+
+## Element similarity in high-dimensional materials representations
+
+The folder `element_similarity` contains the scripts used to produce the results in the Publication: "Element similarity in high-dimensional materials representations"
+
+Installing `SMACT==2.5.3` and `ElementEmbeddings==0.4` via pip should install all the necessary packages required to reproduce the results. We do provide all the packages which were installed at the time of running the notebooks for full transparency in the file `requirements_publication.txt`.

Publication/element_similarity/README.md

@@ -0,0 +1,6 @@
+# Element similarity in high-dimensional materials representations
+
+* The notebook `similarity_periodic_trends/similarity_measures.ipynb` is used to reproduce the dis/similarity heatmaps and 2D projections
+* The notebook `csp/data_radius_ratio_rules.ipynb` is used to get the dataset and calculate the radius ratio rules.
+* The notebook `csp/cosine_embeddings.ipynb` is used to generate the pairwise cosine similarity values
+* The notebook `csp/smact_sp` is used to perform the structure prediction using SMACT

Publication/SI/element_similairty_SI.ipynb → Publication/element_similarity/SI/element_similairty_SI.ipynb

@@ -22,7 +22,10 @@
     "import os\n",
     "import seaborn as sns\n",
     "\n",
-    "sns.set_context(\"paper\", font_scale=1.5)"
+    "sns.set_context(\"paper\", font_scale=1.5)\n",
+    "random_state = 42\n",
+    "reducer_params = {\"random_state\": random_state}\n",
+    "scatter_params = {\"s\": 80}"
    ]
   },
   {
@@ -31,17 +34,24 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "# Set up the 7 embeddings\n",
+    "# Load the embeddings\n",
     "cbfvs = [\n",
-    "    \"magpie_sc\",\n",
+    "    \"magpie\",\n",
+    "    \"matscholar\",\n",
     "    \"mat2vec\",\n",
     "    \"megnet16\",\n",
+    "    \"oliynyk\",\n",
     "    \"random_200\",\n",
-    "    \"matscholar\",\n",
-    "    \"oliynyk_sc\",\n",
     "    \"skipatom\",\n",
     "]\n",
-    "element_embedddings = {cbfv: Embedding.load_data(cbfv) for cbfv in cbfvs}"
+    "element_embedddings = {cbfv: Embedding.load_data(cbfv) for cbfv in cbfvs}\n",
+    "\n",
+    "# Standardise\n",
+    "for embedding in element_embedddings.values():\n",
+    "    print(f\"Attempting to standardise {embedding.embedding_name}...\")\n",
+    "    print(f\" Already standardised: {embedding.is_standardised}\")\n",
+    "    embedding.standardise(inplace=True)\n",
+    "    print(f\"Now standardised: {embedding.is_standardised}\")"
    ]
   },
   {
@@ -79,6 +89,18 @@
     "# del element_embedddings[\"skipatom\"].embeddings[\"Kr\"]"
    ]
   },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Which elements are missing for skipatom\n",
+    "set(element_embedddings[\"magpie\"].element_list) - set(\n",
+    "    element_embedddings[\"skipatom\"].element_list\n",
+    ")"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -109,6 +131,7 @@
     "        show_axislabels=False,\n",
     "        ax=ax,\n",
     "    )\n",
+    "    print(cbfv.embedding_name)\n",
     "    # plt.subplots_adjust(wspace=0.001)\n",
     "axes[-1][-1].remove()\n",
     "\n",
@@ -145,6 +168,70 @@
     "fig.show()"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Kr and Xe contribute to the distorted images for Skipatom."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "element_embedddings[\"skipatom_no_nobles\"] = Embedding.load_data(\"skipatom\")\n",
+    "\n",
+    "for el in [\"Xe\", \"Kr\"]:\n",
+    "    del element_embedddings[\"skipatom_no_nobles\"].embeddings[el]\n",
+    "element_embedddings[\"skipatom_no_nobles\"].standardise(inplace=True)\n",
+    "element_embedddings[\"skipatom_no_nobles\"].embedding_name = \"skipatom (Xe,Kr removed)\""
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "fig, ax = plt.subplots(figsize=(20, 20))\n",
+    "heatmap_plotter(\n",
+    "    embedding=element_embedddings[\"skipatom_no_nobles\"],\n",
+    "    metric=\"euclidean\",\n",
+    "    sortaxisby=\"atomic_number\",\n",
+    "    show_axislabels=True,\n",
+    "    ax=ax,\n",
+    ")\n",
+    "\n",
+    "fig.show()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "fig, (axes) = plt.subplots(4, 2, figsize=(20, 20))\n",
+    "\n",
+    "for ax, cbfv in zip(axes.flatten(), element_embedddings.values()):\n",
+    "    heatmap_plotter(\n",
+    "        embedding=cbfv,\n",
+    "        metric=\"euclidean\",\n",
+    "        sortaxisby=\"atomic_number\",\n",
+    "        show_axislabels=False,\n",
+    "        ax=ax,\n",
+    "    )\n",
+    "    print(cbfv.embedding_name)\n",
+    "    # plt.subplots_adjust(wspace=0.001)\n",
+    "\n",
+    "\n",
+    "fig.tight_layout()\n",
+    "fig.savefig(\"SI_euclidean.pdf\", bbox_inches=\"tight\")\n",
+    "fig.show()"
+   ]
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -169,7 +256,6 @@
     "        ax=ax,\n",
     "    )\n",
     "    # plt.subplots_adjust(wspace=0.001)\n",
-    "axes[-1][-1].remove()\n",
     "\n",
     "\n",
     "fig.tight_layout()\n",
@@ -201,7 +287,6 @@
     "        ax=ax,\n",
     "    )\n",
     "    # plt.subplots_adjust(wspace=0.001)\n",
-    "axes[-1][-1].remove()\n",
     "\n",
     "\n",
     "fig.tight_layout()\n",
@@ -233,7 +318,6 @@
     "        ax=ax,\n",
     "    )\n",
     "    # plt.subplots_adjust(wspace=0.001)\n",
-    "axes[-1][-1].remove()\n",
     "\n",
     "\n",
     "fig.tight_layout()\n",
@@ -265,7 +349,6 @@
     "        ax=ax,\n",
     "    )\n",
     "    # plt.subplots_adjust(wspace=0.001)\n",
-    "axes[-1][-1].remove()\n",
     "\n",
     "\n",
     "fig.tight_layout()\n",
@@ -299,7 +382,6 @@
     "        **heatmap_params\n",
     "    )\n",
     "    # plt.subplots_adjust(wspace=0.001)\n",
-    "axes[-1][-1].remove()\n",
     "\n",
     "\n",
     "fig.tight_layout()\n",
@@ -333,7 +415,6 @@
     "        **heatmap_params\n",
     "    )\n",
     "    # plt.subplots_adjust(wspace=0.001)\n",
-    "axes[-1][-1].remove()\n",
     "\n",
     "\n",
     "fig.tight_layout()\n",
@@ -367,7 +448,6 @@
     "        **heatmap_params\n",
     "    )\n",
     "    # plt.subplots_adjust(wspace=0.001)\n",
-    "axes[-1][-1].remove()\n",
     "\n",
     "\n",
     "fig.tight_layout()\n",
@@ -408,9 +488,11 @@
     "        n_components=2,\n",
     "        ax=ax,\n",
     "        adjusttext=True,\n",
+    "        reducer_params=reducer_params,\n",
+    "        scatter_params=scatter_params,\n",
     "    )\n",
     "    ax.legend().remove()\n",
-    "axes[-1][-1].remove()\n",
+    "\n",
     "\n",
     "handles, labels = ax.get_legend_handles_labels()\n",
     "fig.legend(handles, labels, bbox_to_anchor=(0.54, 1.06), loc=\"upper center\", ncol=5)\n",
@@ -445,9 +527,11 @@
     "        n_components=2,\n",
     "        ax=ax,\n",
     "        # adjusttext=True,\n",
+    "        reducer_params=reducer_params,\n",
+    "        scatter_params=scatter_params,\n",
     "    )\n",
     "    ax.legend().remove()\n",
-    "axes[-1][-1].remove()\n",
+    "\n",
     "\n",
     "handles, labels = ax.get_legend_handles_labels()\n",
     "fig.legend(handles, labels, bbox_to_anchor=(0.54, 1.06), loc=\"upper center\", ncol=5)\n",
@@ -482,9 +566,11 @@
     "        n_components=2,\n",
     "        ax=ax,\n",
     "        adjusttext=True,\n",
+    "        reducer_params=reducer_params,\n",
+    "        scatter_params=scatter_params,\n",
     "    )\n",
     "    ax.legend().remove()\n",
-    "axes[-1][-1].remove()\n",
+    "\n",
     "\n",
     "handles, labels = ax.get_legend_handles_labels()\n",
     "fig.legend(handles, labels, bbox_to_anchor=(0.54, 1.06), loc=\"upper center\", ncol=5)\n",
@@ -538,7 +624,7 @@
     "    ax.set_xlabel(\"Pearson correlation\")\n",
     "    ax.set_ylabel(\"Count\")\n",
     "\n",
-    "axes[-1][-1].remove()\n",
+    "\n",
     "plt.tight_layout()\n",
     "plt.savefig(\"SI_pearson_distribution.pdf\")"
    ]
@@ -566,7 +652,7 @@
     "    ax.set_xlabel(\"Cosine similarity\")\n",
     "    ax.set_ylabel(\"Count\")\n",
     "\n",
-    "axes[-1][-1].remove()\n",
+    "\n",
     "plt.tight_layout()\n",
     "plt.savefig(\"SI_cosine_similarity_distribution.pdf\")"
    ]