Upload hunflair_test.ipynb

hunflair_test.ipynb

@@ -0,0 +1,318 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "C:\\Users\\ranajoy.bhattacharya\\Anaconda3\\lib\\site-packages\\gensim\\similarities\\__init__.py:15: UserWarning: The gensim.similarities.levenshtein submodule is disabled, because the optional Levenshtein package <https://pypi.org/project/python-Levenshtein/> is unavailable. Install Levenhstein (e.g. `pip install python-Levenshtein`) to suppress this warning.\n",
+      "  warnings.warn(msg)\n"
+     ]
+    }
+   ],
+   "source": [
+    "from flair.data import Sentence\n",
+    "from flair.models import MultiTagger"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from flair.tokenization import SciSpacyTokenizer"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "sentence = Sentence(\"Behavioral abnormalities in the Fmr1 KO2 Mouse Model of Fragile X Syndrome\",\n",
+    "                    use_tokenizer=SciSpacyTokenizer())\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "2022-10-27 15:38:14,268 https://nlp.informatik.hu-berlin.de/resources/models/hunflair_smallish_models/cellline/hunflair-celline-v1.0.pt not found in cache, downloading to C:\\Users\\RANAJO~1.BHA\\AppData\\Local\\Temp\\tmpsn3vv9l1\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████████████████████████████████████████████████████████| 1104886372/1104886372 [02:45<00:00, 6695019.03B/s]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "2022-10-27 15:40:59,814 copying C:\\Users\\RANAJO~1.BHA\\AppData\\Local\\Temp\\tmpsn3vv9l1 to cache at C:\\Users\\ranajoy.bhattacharya\\.flair\\models\\hunflair-celline-v1.0.pt\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "2022-10-27 15:41:02,302 removing temp file C:\\Users\\RANAJO~1.BHA\\AppData\\Local\\Temp\\tmpsn3vv9l1\n",
+      "2022-10-27 15:41:02,536 loading file C:\\Users\\ranajoy.bhattacharya\\.flair\\models\\hunflair-celline-v1.0.pt\n",
+      "2022-10-27 15:41:08,505 SequenceTagger predicts: Dictionary with 8 tags: <unk>, O, S-CellLine, B-CellLine, I-CellLine, E-CellLine, <START>, <STOP>\n",
+      "2022-10-27 15:41:09,549 https://nlp.informatik.hu-berlin.de/resources/models/hunflair_allcorpus_models/huner-chemical/hunflair-chemical-full-v1.0.pt not found in cache, downloading to C:\\Users\\RANAJO~1.BHA\\AppData\\Local\\Temp\\tmpw4ze4yrc\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|██████████████████████████████████████████████████████████████| 1104888221/1104888221 [02:06<00:00, 8735803.50B/s]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "2022-10-27 15:43:16,584 copying C:\\Users\\RANAJO~1.BHA\\AppData\\Local\\Temp\\tmpw4ze4yrc to cache at C:\\Users\\ranajoy.bhattacharya\\.flair\\models\\hunflair-chemical-full-v1.0.pt\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "2022-10-27 15:43:18,094 removing temp file C:\\Users\\RANAJO~1.BHA\\AppData\\Local\\Temp\\tmpw4ze4yrc\n",
+      "2022-10-27 15:43:18,208 loading file C:\\Users\\ranajoy.bhattacharya\\.flair\\models\\hunflair-chemical-full-v1.0.pt\n",
+      "2022-10-27 15:43:22,887 SequenceTagger predicts: Dictionary with 8 tags: <unk>, O, S-Chemical, B-Chemical, I-Chemical, E-Chemical, <START>, <STOP>\n",
+      "2022-10-27 15:43:23,826 https://nlp.informatik.hu-berlin.de/resources/models/hunflair_allcorpus_models/huner-disease/hunflair-disease-full-v1.0.pt not found in cache, downloading to C:\\Users\\RANAJO~1.BHA\\AppData\\Local\\Temp\\tmpy18ixxme\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|█████████████████████████████████████████████████████████████| 1104886193/1104886193 [01:33<00:00, 11804678.71B/s]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "2022-10-27 15:44:57,965 copying C:\\Users\\RANAJO~1.BHA\\AppData\\Local\\Temp\\tmpy18ixxme to cache at C:\\Users\\ranajoy.bhattacharya\\.flair\\models\\hunflair-disease-full-v1.0.pt\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "2022-10-27 15:44:59,866 removing temp file C:\\Users\\RANAJO~1.BHA\\AppData\\Local\\Temp\\tmpy18ixxme\n",
+      "2022-10-27 15:45:00,019 loading file C:\\Users\\ranajoy.bhattacharya\\.flair\\models\\hunflair-disease-full-v1.0.pt\n",
+      "2022-10-27 15:45:06,491 SequenceTagger predicts: Dictionary with 8 tags: <unk>, O, B-Disease, E-Disease, I-Disease, S-Disease, <START>, <STOP>\n",
+      "2022-10-27 15:45:07,641 https://nlp.informatik.hu-berlin.de/resources/models/hunflair_allcorpus_models/huner-gene/hunflair-gene-full-v1.0.pt not found in cache, downloading to C:\\Users\\RANAJO~1.BHA\\AppData\\Local\\Temp\\tmp1qgb402b\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|█████████████████████████████████████████████████████████████| 1104887211/1104887211 [01:22<00:00, 13387243.47B/s]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "2022-10-27 15:46:30,726 copying C:\\Users\\RANAJO~1.BHA\\AppData\\Local\\Temp\\tmp1qgb402b to cache at C:\\Users\\ranajoy.bhattacharya\\.flair\\models\\hunflair-gene-full-v1.0.pt\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "2022-10-27 15:46:32,696 removing temp file C:\\Users\\RANAJO~1.BHA\\AppData\\Local\\Temp\\tmp1qgb402b\n",
+      "2022-10-27 15:46:32,831 loading file C:\\Users\\ranajoy.bhattacharya\\.flair\\models\\hunflair-gene-full-v1.0.pt\n",
+      "2022-10-27 15:46:38,569 SequenceTagger predicts: Dictionary with 8 tags: <unk>, O, S-Gene, B-Gene, I-Gene, E-Gene, <START>, <STOP>\n",
+      "2022-10-27 15:46:39,870 https://nlp.informatik.hu-berlin.de/resources/models/hunflair_allcorpus_models/huner-species/hunflair-species-full-v1.1.pt not found in cache, downloading to C:\\Users\\RANAJO~1.BHA\\AppData\\Local\\Temp\\tmpj1y4y_6v\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "100%|█████████████████████████████████████████████████████████████| 1104886931/1104886931 [01:16<00:00, 14508392.59B/s]"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "2022-10-27 15:47:56,571 copying C:\\Users\\RANAJO~1.BHA\\AppData\\Local\\Temp\\tmpj1y4y_6v to cache at C:\\Users\\ranajoy.bhattacharya\\.flair\\models\\hunflair-species-full-v1.1.pt\n"
+     ]
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "\n"
+     ]
+    },
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "2022-10-27 15:47:58,616 removing temp file C:\\Users\\RANAJO~1.BHA\\AppData\\Local\\Temp\\tmpj1y4y_6v\n",
+      "2022-10-27 15:47:58,765 loading file C:\\Users\\ranajoy.bhattacharya\\.flair\\models\\hunflair-species-full-v1.1.pt\n",
+      "2022-10-27 15:48:04,720 SequenceTagger predicts: Dictionary with 8 tags: <unk>, O, S-Species, B-Species, I-Species, E-Species, <START>, <STOP>\n"
+     ]
+    }
+   ],
+   "source": [
+    "# load biomedical tagger\n",
+    "tagger = MultiTagger.load(\"hunflair\")\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Span[0:2]: \"Behavioral abnormalities\" → Disease (0.6736)\n",
+      "Span[9:12]: \"Fragile X Syndrome\" → Disease (0.99)\n",
+      "Span[4:5]: \"Fmr1\" → Gene (0.838)\n",
+      "Span[6:7]: \"Mouse\" → Species (0.9979)\n"
+     ]
+    }
+   ],
+   "source": [
+    "for annotation_layer in sentence.annotation_layers.keys():\n",
+    "    for entity in sentence.get_spans(annotation_layer):\n",
+    "        print(entity)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "2022-10-27 18:06:43,204 loading file C:\\Users\\ranajoy.bhattacharya\\.flair\\models\\hunflair-chemical-full-v1.0.pt\n",
+      "2022-10-27 18:06:49,684 SequenceTagger predicts: Dictionary with 8 tags: <unk>, O, S-Chemical, B-Chemical, I-Chemical, E-Chemical, <START>, <STOP>\n"
+     ]
+    }
+   ],
+   "source": [
+    "sentence = Sentence(\"Polycrystalline Li4+MxSi1-xO4 (M = B, Al) samples were prepared from reagent grade H3BO3 Al2O3, SiO2 and either LiOH*H2O (H samples) or Li2CO3 (C samples). Intimate stoichiometric mixtures were placed in an Al2O3 crucible or an Au foil boat and fired.\")\n",
+    "\n",
+    "# load biomedical tagger\n",
+    "tagger = MultiTagger.load(\"hunflair-chemical\")\n",
+    "\n",
+    "# tag sentence\n",
+    "tagger.predict(sentence)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 13,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Span[1:2]: \"Li4\" → Chemical (0.754)\n",
+      "Span[3:4]: \"MxSi1-xO4\" → Chemical (0.8234)\n",
+      "Span[9:10]: \"Al\" → Chemical (0.9942)\n",
+      "Span[17:18]: \"H3BO3\" → Chemical (0.7564)\n",
+      "Span[18:19]: \"Al2O3\" → Chemical (0.5558)\n",
+      "Span[20:21]: \"SiO2\" → Chemical (0.973)\n",
+      "Span[23:24]: \"LiOH\" → Chemical (0.9721)\n",
+      "Span[25:26]: \"H2O\" → Chemical (0.8947)\n",
+      "Span[31:32]: \"Li2CO3\" → Chemical (0.9224)\n",
+      "Span[44:45]: \"Al2O3\" → Chemical (0.9847)\n",
+      "Span[48:49]: \"Au\" → Chemical (0.9787)\n"
+     ]
+    }
+   ],
+   "source": [
+    "for annotation_layer in sentence.annotation_layers.keys():\n",
+    "    for entity in sentence.get_spans(annotation_layer):\n",
+    "        print(entity)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.7.6"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}