fig1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "1530ce28-a0ea-4e96-8b2f-71b900cb3049",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true
   },
   "source": [
    "## Data Preprocess"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "f55eb43e-cc6b-4b3f-aa63-0179ffd0c469",
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "import os\n",
    "os.environ['CUDA_VISIBLE_DEVICES'] = '9'\n",
    "from deepsramp import *"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1b440ec5-f3f4-42e0-bbd9-2cab298e15ca",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true,
    "tags": []
   },
   "source": [
    "### Proprocess GTF and FASTA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "7ee6fa44-8076-48bd-a445-64209cca1e6f",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/nfs/fanrui/work/deepsramp/deepsramp/preprocess.py:30: DtypeWarning: Columns (0) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "  gtf = pd.read_csv(filename, sep='\\t', comment='#', header=None)\n",
      "51662809it [02:01, 425761.41it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1min 34s, sys: 1min 3s, total: 2min 38s\n",
      "Wall time: 2min 35s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "gtf = read_gtf('data/Homo_sapiens.GRCh38.79.gtf.gz')\n",
    "seqs = read_seq('data/Homo_sapiens.GRCh38.dna.primary_assembly.fa.gz')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "61bfc299-119f-4c0d-a363-2fdf6066a317",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 15.1 s, sys: 564 ms, total: 15.6 s\n",
      "Wall time: 15.6 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "sdf = get_mature_splice(gtf, seqs, grp=['id', 'trans'], keepmax=False)\n",
    "ndf = get_mature_cds(gtf, sdf, grp=['id', 'trans'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "4448a7e0-829f-4e2c-b482-ad91c60e8926",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/nfs/fanrui/work/deepsramp/deepsramp/preprocess.py:30: DtypeWarning: Columns (0) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "  gtf = pd.read_csv(filename, sep='\\t', comment='#', header=None)\n",
      "45470460it [01:34, 479865.13it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1min 21s, sys: 39.7 s, total: 2min 1s\n",
      "Wall time: 2min\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "gtfm = read_gtf('data/Mus_musculus.GRCm39.109.gtf.gz')\n",
    "seqsm = read_seq('data/Mus_musculus.GRCm39.dna.primary_assembly.fa.gz')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "59b36dd6-a5ed-47f3-b3d1-05b7f5091dd8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 11 s, sys: 735 ms, total: 11.7 s\n",
      "Wall time: 11.7 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "sdfm = get_mature_splice(gtfm, seqsm, grp=['id', 'trans'], keepmax=False)\n",
    "ndfm = get_mature_cds(gtfm, sdfm, grp=['id', 'trans'])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "31f090be-0fc0-47fb-a5e4-db55d2c0af1d",
   "metadata": {},
   "source": [
    "#### Full Trans Mode"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "aee7d236-f172-4c2e-ac0d-8f73d65e47e5",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "sdf = get_mature_splice(gtf, seqs, grp=['id', 'trans'], keepmax=False)\n",
    "ndf = get_full_splice(sdf, seqs)\n",
    "ndf = get_full_cds(gtf, ndf, grp=['id', 'trans'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "53d9c75f-6242-4a79-a7d7-cd859a714661",
   "metadata": {},
   "outputs": [],
   "source": [
    "sdfm = get_mature_splice(gtfm, seqsm, grp=['id', 'trans'], keepmax=False)\n",
    "ndfm = get_full_splice(sdfm, seqsm)\n",
    "ndfm = get_full_cds(gtfm, ndfm, grp=['id', 'trans'])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "475b7d8e-0e25-4492-9f9a-eea847f56294",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true
   },
   "source": [
    "### Preprocess of SRAMP1 data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "8551c7d3-278e-4388-b35f-639e8e0b3ded",
   "metadata": {
    "tags": []
   },
   "outputs": [],
   "source": [
    "def get_sramp_dataset_all(file, rndf, downsample=False, mode='mature'):\n",
    "    train = pd.read_csv(file)\n",
    "    train['Class'] = (train['Class'] + 1) // 2\n",
    "    train['Position'] -= 1\n",
    "    keep = []\n",
    "    for _, trans, pos, claz in tqdm(train.itertuples(), total=train.shape[0]):\n",
    "        if trans not in rndf.index: \n",
    "            keep += [0]\n",
    "        elif DRACH_PAT.findall(rndf.loc[trans].seq[pos-2:pos+3]):\n",
    "            keep += [1]\n",
    "        else:\n",
    "            keep += [0]\n",
    "\n",
    "    train['keep'] = keep\n",
    "\n",
    "    train = train[train.keep == 1]\n",
    "    if downsample:\n",
    "        pos = train[train.Class == 1]\n",
    "        neg = train[train.Class == 0].sample(pos.shape[0])\n",
    "        train = pd.concat([pos, neg], axis=0)\n",
    "\n",
    "    lookupdf = rndf.reset_index().set_index('id')\n",
    "    res = []\n",
    "    for idx, trans, pos, claz, _ in tqdm(train.itertuples(), total=train.shape[0]):\n",
    "        tdf = rndf.loc[trans]\n",
    "        genome_pos = get_genome_pos(pos, tdf, mode=mode)\n",
    "        tdf = lookupdf.loc[[tdf.id]].copy()\n",
    "        tres = []\n",
    "        for _, ttdf in tdf.iterrows() :\n",
    "            if mode == 'mature':\n",
    "                trans_pos = get_mature_pos(genome_pos, ttdf)\n",
    "            else:\n",
    "                trans_pos = get_full_pos(genome_pos, ttdf)\n",
    "            tres += [trans_pos]\n",
    "        tdf['pos'] = tres\n",
    "        tdf = tdf[tdf.pos != -1]\n",
    "        tdf['label'] = claz\n",
    "        tdf['grp'] = tdf[0] + tdf[6] + str(genome_pos)\n",
    "        # tdf['grp'] = file + str(idx)\n",
    "        res += [tdf.dropna()]\n",
    "    res = pd.concat(res, axis=0)\n",
    "\n",
    "    return res"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "c4992aef-f806-4852-bf64-747a7dfdc9d0",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 516890/516890 [00:16<00:00, 31785.41it/s] \n",
      "100%|██████████| 51088/51088 [02:04<00:00, 410.20it/s]\n",
      "100%|██████████| 133826/133826 [00:04<00:00, 32809.36it/s]\n",
      "100%|██████████| 73551/73551 [02:58<00:00, 411.28it/s]\n"
     ]
    }
   ],
   "source": [
    "rndf = ndf.reset_index().set_index('trans')\n",
    "traindf = get_sramp_dataset_all('data/sramp1/mature_train.csv', rndf, downsample=True)\n",
    "testdf = get_sramp_dataset_all('data/sramp1/mature_test.csv', rndf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "3771cb0c-e2f7-4cce-aab6-33d8f2777d72",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 516890/516890 [00:11<00:00, 43958.42it/s] \n",
      "100%|██████████| 32528/32528 [01:15<00:00, 429.94it/s]\n",
      "100%|██████████| 133826/133826 [00:03<00:00, 42536.23it/s]\n",
      "100%|██████████| 45690/45690 [01:47<00:00, 424.79it/s]\n"
     ]
    }
   ],
   "source": [
    "rndfm = ndfm.reset_index().set_index('trans')\n",
    "traindfm = get_sramp_dataset_all('data/sramp1/mature_train.csv', rndfm, downsample=True)\n",
    "testdfm = get_sramp_dataset_all('data/sramp1/mature_test.csv', rndfm)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "462c773c-cffa-41f6-9a6f-dd1903a68fd9",
   "metadata": {},
   "outputs": [],
   "source": [
    "traindf = pd.concat([traindf, traindfm])\n",
    "testdf = pd.concat([testdf, testdfm])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "7bb9e0f3-70da-4904-94e0-dc052a63c478",
   "metadata": {},
   "outputs": [],
   "source": [
    "utils.save([traindf, testdf], 'data/sramp1/sramp1_mature.data')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ad2a7be3-a1b3-456f-a952-55bf70f049a8",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "fe7b95ef-c93c-46e5-ad25-5e3b134b4910",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 612744/612744 [00:18<00:00, 33172.43it/s] \n",
      "100%|██████████| 56920/56920 [02:20<00:00, 404.23it/s]\n",
      "100%|██████████| 157729/157729 [00:04<00:00, 32672.97it/s]\n",
      "100%|██████████| 83343/83343 [03:23<00:00, 410.34it/s]\n"
     ]
    }
   ],
   "source": [
    "rndf = ndf.reset_index().set_index('trans')\n",
    "traindf = get_sramp_dataset_all('data/sramp1/full_train.csv', rndf, downsample=True, mode='full')\n",
    "testdf = get_sramp_dataset_all('data/sramp1/full_test.csv', rndf, mode='full')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "8fc915a7-baf8-4bb3-b230-9a848b74e596",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 612744/612744 [00:12<00:00, 48865.11it/s] \n",
      "100%|██████████| 36258/36258 [01:24<00:00, 430.65it/s]\n",
      "100%|██████████| 157729/157729 [00:03<00:00, 49429.23it/s]\n",
      "100%|██████████| 49395/49395 [01:56<00:00, 425.32it/s]\n"
     ]
    }
   ],
   "source": [
    "rndfm = ndfm.reset_index().set_index('trans')\n",
    "traindfm = get_sramp_dataset_all('data/sramp1/full_train.csv', rndfm, downsample=True, mode='full')\n",
    "testdfm = get_sramp_dataset_all('data/sramp1/full_test.csv', rndfm, mode='full')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "ba4b9f6e-85ed-43c1-a4ff-abdb23c1a7c3",
   "metadata": {},
   "outputs": [],
   "source": [
    "traindf = pd.concat([traindf, traindfm])\n",
    "testdf = pd.concat([testdf, testdfm])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "5f6b9aa7-47a5-47f4-a011-95a3e5b8d33d",
   "metadata": {},
   "outputs": [],
   "source": [
    "utils.save([traindf, testdf], 'data/sramp1/sramp1_full.data')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "104347ac-6f28-4538-b07e-6a3cbe7627c3",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "cf745ee7-48b6-4cbd-9afa-986fe267a51a",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true
   },
   "source": [
    "## Train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "58eebe09-b531-42ab-a104-513abba2ebe3",
   "metadata": {},
   "outputs": [],
   "source": [
    "traindf, testdf = utils.load('data/sramp1/sramp1_mature.data')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "134d12f7-edf1-4e4e-918b-2c2e18c3dd59",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6195121365921553"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(lambda x: x[x>1].shape[0] / x.shape[0])(traindf.grp.value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "e7ea4c17-56cc-44a1-bb38-9f3c52982981",
   "metadata": {},
   "outputs": [],
   "source": [
    "traindf, testdf = utils.load('data/sramp1/sramp1_full.data')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "ea4b2cfa-09cd-44fb-8dd7-41b6e0ccc3ee",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 134654/134654 [01:42<00:00, 1310.48it/s]\n",
      "100%|██████████| 194716/194716 [02:09<00:00, 1508.07it/s]\n"
     ]
    }
   ],
   "source": [
    "trainds = df2ds_multi(traindf)\n",
    "traindl = ds2dl(trainds, batch_size=128, drop_last=True, num_workers=4)\n",
    "\n",
    "testds = df2ds_multi(testdf)\n",
    "testdl = ds2dl(testds, batch_size=128, shuffle=False, num_workers=4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "59d14e27-ba4e-4aa0-b8dc-d81f45617b49",
   "metadata": {},
   "outputs": [],
   "source": [
    "import time\n",
    "\n",
    "# print(f'Building model')\n",
    "device = \"cuda\" if torch.cuda.is_available() else \"cpu\"\n",
    "\n",
    "test_res = []\n",
    "for i in ['genomeonly', 'seqonly', 'full']:\n",
    "    utils.setup_seed(42)\n",
    "    model = MultiSRAMP(mode=i).to(device)\n",
    "    loss_fn = TriLoss()\n",
    "    optimizer = torch.optim.Adam(model.parameters(), lr=1e-3)\n",
    "    scheduler = torch.optim.lr_scheduler.CosineAnnealingWarmRestarts(optimizer, 1, 2)\n",
    "\n",
    "    print(f'Start to train')\n",
    "    start = time.time()\n",
    "    train(model, loss_fn, optimizer, scheduler, traindl, testdl, epochs=31, device=device)\n",
    "    end = time.time()\n",
    "    test_res += [test_loop(testdl, model, device, loss_fn)]\n",
    "    print(end-start)\n",
    "    \n",
    "    torch.save(model, f'model/{i}_400_mature.model')\n",
    "    utils.save(test_res, f'model/{i}_400_mature.metrics')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3078e9f0-cba2-4ed8-9968-a400e3fd2cf6",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true
   },
   "source": [
    "## Time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "521b1c13-4e33-4f10-99ef-13260768fed7",
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Start to train\n",
      "[Epoch 01] loss: 0.697706  [     0/ 50884] \n",
      "[Epoch 02] loss: 0.397413  [     0/ 50884] \n",
      "[Epoch 03] loss: 0.325380  [     0/ 50884] \n",
      "[Epoch 04] loss: 0.317262  [     0/ 50884] \n",
      "[Epoch 05] loss: 0.325515  [     0/ 50884] \n",
      "[Epoch 06] loss: 0.263835  [     0/ 50884] \n",
      "[Epoch 07] loss: 0.331850  [     0/ 50884] \n",
      "[Epoch 08] loss: 0.281225  [     0/ 50884] \n",
      "[Epoch 09] loss: 0.318910  [     0/ 50884] \n",
      "[Epoch 10] loss: 0.328378  [     0/ 50884] \n",
      "[Epoch 11] loss: 0.334452  [     0/ 50884] \n",
      "[Epoch 12] loss: 0.239673  [     0/ 50884] \n",
      "[Epoch 13] loss: 0.315308  [     0/ 50884] \n",
      "[Epoch 14] loss: 0.305838  [     0/ 50884] \n",
      "[Epoch 15] loss: 0.305151  [     0/ 50884] \n",
      "[Epoch 16] loss: 0.298173  [     0/ 50884] \n",
      "[Epoch 17] loss: 0.250777  [     0/ 50884] \n",
      "[Epoch 18] loss: 0.334884  [     0/ 50884] \n",
      "[Epoch 19] loss: 0.255954  [     0/ 50884] \n",
      "[Epoch 20] loss: 0.219362  [     0/ 50884] \n",
      "[Epoch 21] loss: 0.331875  [     0/ 50884] \n",
      "[Epoch 22] loss: 0.290480  [     0/ 50884] \n",
      "[Epoch 23] loss: 0.259607  [     0/ 50884] \n",
      "[Epoch 24] loss: 0.243387  [     0/ 50884] \n",
      "[Epoch 25] loss: 0.212428  [     0/ 50884] \n",
      "[Epoch 26] loss: 0.255065  [     0/ 50884] \n",
      "[Epoch 27] loss: 0.274836  [     0/ 50884] \n",
      "[Epoch 28] loss: 0.232988  [     0/ 50884] \n",
      "[Epoch 29] loss: 0.260712  [     0/ 50884] \n",
      "[Epoch 30] loss: 0.211421  [     0/ 50884] \n",
      "[Epoch 31] loss: 0.081535  [     0/ 50884] \n",
      "[Test] ACC: 83.6%, Loss: 0.371845, AUROC: 0.9393, AUPRC: 0.5948, Size: 73484, Ratio: 0.08942\n",
      "3936.254439353943\n",
      "Start to train\n",
      "[Epoch 01] loss: 0.695828  [     0/ 50884] \n",
      "[Epoch 02] loss: 0.412644  [     0/ 50884] \n",
      "[Epoch 03] loss: 0.344995  [     0/ 50884] \n",
      "[Epoch 04] loss: 0.301291  [     0/ 50884] \n",
      "[Epoch 05] loss: 0.330993  [     0/ 50884] \n",
      "[Epoch 06] loss: 0.278263  [     0/ 50884] \n",
      "[Epoch 07] loss: 0.352893  [     0/ 50884] \n",
      "[Epoch 08] loss: 0.269503  [     0/ 50884] \n",
      "[Epoch 09] loss: 0.305950  [     0/ 50884] \n",
      "[Epoch 10] loss: 0.322306  [     0/ 50884] \n",
      "[Epoch 11] loss: 0.334365  [     0/ 50884] \n",
      "[Epoch 12] loss: 0.234415  [     0/ 50884] \n",
      "[Epoch 13] loss: 0.320165  [     0/ 50884] \n",
      "[Epoch 14] loss: 0.295706  [     0/ 50884] \n",
      "[Epoch 15] loss: 0.305872  [     0/ 50884] \n",
      "[Epoch 16] loss: 0.292245  [     0/ 50884] \n",
      "[Epoch 17] loss: 0.274851  [     0/ 50884] \n",
      "[Epoch 18] loss: 0.327472  [     0/ 50884] \n",
      "[Epoch 19] loss: 0.244147  [     0/ 50884] \n",
      "[Epoch 20] loss: 0.232697  [     0/ 50884] \n",
      "[Epoch 21] loss: 0.275393  [     0/ 50884] \n",
      "[Epoch 22] loss: 0.273809  [     0/ 50884] \n",
      "[Epoch 23] loss: 0.238268  [     0/ 50884] \n",
      "[Epoch 24] loss: 0.265136  [     0/ 50884] \n",
      "[Epoch 25] loss: 0.212930  [     0/ 50884] \n",
      "[Epoch 26] loss: 0.260078  [     0/ 50884] \n",
      "[Epoch 27] loss: 0.292316  [     0/ 50884] \n",
      "[Epoch 28] loss: 0.259286  [     0/ 50884] \n",
      "[Epoch 29] loss: 0.239312  [     0/ 50884] \n",
      "[Epoch 30] loss: 0.227728  [     0/ 50884] \n",
      "[Epoch 31] loss: 0.087849  [     0/ 50884] \n",
      "[Test] ACC: 84.2%, Loss: 0.327270, AUROC: 0.9446, AUPRC: 0.6252, Size: 73484, Ratio: 0.08942\n",
      "3449.1340901851654\n",
      "Start to train\n",
      "[Epoch 01] loss: 0.699485  [     0/ 50884] \n",
      "[Epoch 02] loss: 0.398157  [     0/ 50884] \n",
      "[Epoch 03] loss: 0.316649  [     0/ 50884] \n",
      "[Epoch 04] loss: 0.264297  [     0/ 50884] \n",
      "[Epoch 05] loss: 0.315859  [     0/ 50884] \n",
      "[Epoch 06] loss: 0.268343  [     0/ 50884] \n",
      "[Epoch 07] loss: 0.347523  [     0/ 50884] \n",
      "[Epoch 08] loss: 0.258351  [     0/ 50884] \n",
      "[Epoch 09] loss: 0.326435  [     0/ 50884] \n",
      "[Epoch 10] loss: 0.321688  [     0/ 50884] \n",
      "[Epoch 11] loss: 0.322513  [     0/ 50884] \n",
      "[Epoch 12] loss: 0.235992  [     0/ 50884] \n",
      "[Epoch 13] loss: 0.307291  [     0/ 50884] \n",
      "[Epoch 14] loss: 0.290006  [     0/ 50884] \n",
      "[Epoch 15] loss: 0.293347  [     0/ 50884] \n",
      "[Epoch 16] loss: 0.269425  [     0/ 50884] \n",
      "[Epoch 17] loss: 0.237173  [     0/ 50884] \n",
      "[Epoch 18] loss: 0.309021  [     0/ 50884] \n",
      "[Epoch 19] loss: 0.245842  [     0/ 50884] \n",
      "[Epoch 20] loss: 0.234754  [     0/ 50884] \n",
      "[Epoch 21] loss: 0.316549  [     0/ 50884] \n",
      "[Epoch 22] loss: 0.249334  [     0/ 50884] \n",
      "[Epoch 23] loss: 0.224237  [     0/ 50884] \n",
      "[Epoch 24] loss: 0.240383  [     0/ 50884] \n",
      "[Epoch 25] loss: 0.209735  [     0/ 50884] \n",
      "[Epoch 26] loss: 0.240820  [     0/ 50884] \n",
      "[Epoch 27] loss: 0.285406  [     0/ 50884] \n",
      "[Epoch 28] loss: 0.281604  [     0/ 50884] \n",
      "[Epoch 29] loss: 0.240534  [     0/ 50884] \n",
      "[Epoch 30] loss: 0.229935  [     0/ 50884] \n",
      "[Epoch 31] loss: 0.072858  [     0/ 50884] \n",
      "[Test] ACC: 84.6%, Loss: 0.323679, AUROC: 0.9454, AUPRC: 0.6322, Size: 73484, Ratio: 0.08942\n",
      "4817.658853769302\n",
      "Start to train\n",
      "[Epoch 01] loss: 0.694798  [     0/ 50884] \n",
      "[Epoch 02] loss: 0.393669  [     0/ 50884] \n",
      "[Epoch 03] loss: 0.327131  [     0/ 50884] \n",
      "[Epoch 04] loss: 0.258649  [     0/ 50884] \n",
      "[Epoch 05] loss: 0.334181  [     0/ 50884] \n",
      "[Epoch 06] loss: 0.250576  [     0/ 50884] \n",
      "[Epoch 07] loss: 0.333707  [     0/ 50884] \n",
      "[Epoch 08] loss: 0.254289  [     0/ 50884] \n",
      "[Epoch 09] loss: 0.342539  [     0/ 50884] \n",
      "[Epoch 10] loss: 0.288540  [     0/ 50884] \n",
      "[Epoch 11] loss: 0.312249  [     0/ 50884] \n",
      "[Epoch 12] loss: 0.222978  [     0/ 50884] \n",
      "[Epoch 13] loss: 0.294803  [     0/ 50884] \n",
      "[Epoch 14] loss: 0.297717  [     0/ 50884] \n",
      "[Epoch 15] loss: 0.280747  [     0/ 50884] \n",
      "[Epoch 16] loss: 0.271372  [     0/ 50884] \n",
      "[Epoch 17] loss: 0.261383  [     0/ 50884] \n",
      "[Epoch 18] loss: 0.307983  [     0/ 50884] \n",
      "[Epoch 19] loss: 0.224329  [     0/ 50884] \n",
      "[Epoch 20] loss: 0.231732  [     0/ 50884] \n",
      "[Epoch 21] loss: 0.292148  [     0/ 50884] \n",
      "[Epoch 22] loss: 0.260685  [     0/ 50884] \n",
      "[Epoch 23] loss: 0.212500  [     0/ 50884] \n",
      "[Epoch 24] loss: 0.245655  [     0/ 50884] \n",
      "[Epoch 25] loss: 0.189844  [     0/ 50884] \n",
      "[Epoch 26] loss: 0.232033  [     0/ 50884] \n",
      "[Epoch 27] loss: 0.313598  [     0/ 50884] \n",
      "[Epoch 28] loss: 0.222308  [     0/ 50884] \n",
      "[Epoch 29] loss: 0.207179  [     0/ 50884] \n",
      "[Epoch 30] loss: 0.275420  [     0/ 50884] \n",
      "[Epoch 31] loss: 0.073782  [     0/ 50884] \n",
      "[Test] ACC: 84.6%, Loss: 0.323307, AUROC: 0.9473, AUPRC: 0.6397, Size: 73484, Ratio: 0.08942\n",
      "6526.958355665207\n",
      "Start to train\n",
      "[Epoch 01] loss: 0.705053  [     0/ 50884] \n",
      "[Epoch 02] loss: 0.418378  [     0/ 50884] \n",
      "[Epoch 03] loss: 0.298733  [     0/ 50884] \n",
      "[Epoch 04] loss: 0.279483  [     0/ 50884] \n",
      "[Epoch 05] loss: 0.305343  [     0/ 50884] \n",
      "[Epoch 06] loss: 0.250217  [     0/ 50884] \n",
      "[Epoch 07] loss: 0.317781  [     0/ 50884] \n",
      "[Epoch 08] loss: 0.254062  [     0/ 50884] \n",
      "[Epoch 09] loss: 0.337008  [     0/ 50884] \n",
      "[Epoch 10] loss: 0.270969  [     0/ 50884] \n",
      "[Epoch 11] loss: 0.286741  [     0/ 50884] \n",
      "[Epoch 12] loss: 0.219029  [     0/ 50884] \n",
      "[Epoch 13] loss: 0.294231  [     0/ 50884] \n",
      "[Epoch 14] loss: 0.278340  [     0/ 50884] \n",
      "[Epoch 15] loss: 0.288995  [     0/ 50884] \n",
      "[Epoch 16] loss: 0.274254  [     0/ 50884] \n",
      "[Epoch 17] loss: 0.228678  [     0/ 50884] \n",
      "[Epoch 18] loss: 0.280312  [     0/ 50884] \n",
      "[Epoch 19] loss: 0.254831  [     0/ 50884] \n",
      "[Epoch 20] loss: 0.235003  [     0/ 50884] \n",
      "[Epoch 21] loss: 0.270811  [     0/ 50884] \n",
      "[Epoch 22] loss: 0.234100  [     0/ 50884] \n",
      "[Epoch 23] loss: 0.227198  [     0/ 50884] \n",
      "[Epoch 24] loss: 0.264763  [     0/ 50884] \n",
      "[Epoch 25] loss: 0.219600  [     0/ 50884] \n",
      "[Epoch 26] loss: 0.230957  [     0/ 50884] \n",
      "[Epoch 27] loss: 0.267522  [     0/ 50884] \n",
      "[Epoch 28] loss: 0.230350  [     0/ 50884] \n",
      "[Epoch 29] loss: 0.229891  [     0/ 50884] \n",
      "[Epoch 30] loss: 0.257521  [     0/ 50884] \n",
      "[Epoch 31] loss: 0.077211  [     0/ 50884] \n",
      "[Test] ACC: 84.4%, Loss: 0.327458, AUROC: 0.9475, AUPRC: 0.6379, Size: 73484, Ratio: 0.08942\n",
      "8414.479652166367\n",
      "Start to train\n",
      "[Epoch 01] loss: 0.700831  [     0/ 50884] \n",
      "[Epoch 02] loss: 0.397870  [     0/ 50884] \n",
      "[Epoch 03] loss: 0.325856  [     0/ 50884] \n",
      "[Epoch 04] loss: 0.276241  [     0/ 50884] \n",
      "[Epoch 05] loss: 0.327574  [     0/ 50884] \n",
      "[Epoch 06] loss: 0.284779  [     0/ 50884] \n",
      "[Epoch 07] loss: 0.319844  [     0/ 50884] \n",
      "[Epoch 08] loss: 0.248856  [     0/ 50884] \n",
      "[Epoch 09] loss: 0.328809  [     0/ 50884] \n",
      "[Epoch 10] loss: 0.286119  [     0/ 50884] \n",
      "[Epoch 11] loss: 0.296220  [     0/ 50884] \n",
      "[Epoch 12] loss: 0.212526  [     0/ 50884] \n",
      "[Epoch 13] loss: 0.293315  [     0/ 50884] \n",
      "[Epoch 14] loss: 0.284641  [     0/ 50884] \n",
      "[Epoch 15] loss: 0.288499  [     0/ 50884] \n",
      "[Epoch 16] loss: 0.272310  [     0/ 50884] \n",
      "[Epoch 17] loss: 0.253337  [     0/ 50884] \n",
      "[Epoch 18] loss: 0.297556  [     0/ 50884] \n",
      "[Epoch 19] loss: 0.235367  [     0/ 50884] \n",
      "[Epoch 20] loss: 0.213624  [     0/ 50884] \n",
      "[Epoch 21] loss: 0.284325  [     0/ 50884] \n",
      "[Epoch 22] loss: 0.256837  [     0/ 50884] \n",
      "[Epoch 23] loss: 0.229440  [     0/ 50884] \n",
      "[Epoch 24] loss: 0.239254  [     0/ 50884] \n",
      "[Epoch 25] loss: 0.219151  [     0/ 50884] \n",
      "[Epoch 26] loss: 0.221081  [     0/ 50884] \n",
      "[Epoch 27] loss: 0.297195  [     0/ 50884] \n",
      "[Epoch 28] loss: 0.192672  [     0/ 50884] \n",
      "[Epoch 29] loss: 0.201477  [     0/ 50884] \n",
      "[Epoch 30] loss: 0.239892  [     0/ 50884] \n",
      "[Epoch 31] loss: 0.073203  [     0/ 50884] \n",
      "[Test] ACC: 84.9%, Loss: 0.325903, AUROC: 0.9486, AUPRC: 0.6451, Size: 73484, Ratio: 0.08942\n",
      "10551.28130531311\n",
      "Start to train\n",
      "[Epoch 01] loss: 0.697853  [     0/ 50884] \n",
      "[Epoch 02] loss: 0.382298  [     0/ 50884] \n",
      "[Epoch 03] loss: 0.309890  [     0/ 50884] \n",
      "[Epoch 04] loss: 0.275921  [     0/ 50884] \n",
      "[Epoch 05] loss: 0.293817  [     0/ 50884] \n",
      "[Epoch 06] loss: 0.267675  [     0/ 50884] \n",
      "[Epoch 07] loss: 0.321764  [     0/ 50884] \n",
      "[Epoch 08] loss: 0.246260  [     0/ 50884] \n",
      "[Epoch 09] loss: 0.319855  [     0/ 50884] \n",
      "[Epoch 10] loss: 0.258185  [     0/ 50884] \n",
      "[Epoch 11] loss: 0.321655  [     0/ 50884] \n",
      "[Epoch 12] loss: 0.227389  [     0/ 50884] \n",
      "[Epoch 13] loss: 0.294497  [     0/ 50884] \n",
      "[Epoch 14] loss: 0.280324  [     0/ 50884] \n",
      "[Epoch 15] loss: 0.283940  [     0/ 50884] \n",
      "[Epoch 16] loss: 0.269728  [     0/ 50884] \n",
      "[Epoch 17] loss: 0.249657  [     0/ 50884] \n",
      "[Epoch 18] loss: 0.316336  [     0/ 50884] \n",
      "[Epoch 19] loss: 0.234257  [     0/ 50884] \n",
      "[Epoch 20] loss: 0.229422  [     0/ 50884] \n",
      "[Epoch 21] loss: 0.282561  [     0/ 50884] \n",
      "[Epoch 22] loss: 0.262406  [     0/ 50884] \n",
      "[Epoch 23] loss: 0.245291  [     0/ 50884] \n",
      "[Epoch 24] loss: 0.236519  [     0/ 50884] \n",
      "[Epoch 25] loss: 0.226788  [     0/ 50884] \n",
      "[Epoch 26] loss: 0.217545  [     0/ 50884] \n",
      "[Epoch 27] loss: 0.277110  [     0/ 50884] \n",
      "[Epoch 28] loss: 0.217743  [     0/ 50884] \n",
      "[Epoch 29] loss: 0.234915  [     0/ 50884] \n",
      "[Epoch 30] loss: 0.219790  [     0/ 50884] \n",
      "[Epoch 31] loss: 0.069766  [     0/ 50884] \n",
      "[Test] ACC: 84.9%, Loss: 0.323953, AUROC: 0.9475, AUPRC: 0.6377, Size: 73484, Ratio: 0.08942\n",
      "12895.43722820282\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "\n",
    "# print(f'Building model')\n",
    "device = \"cuda\" if torch.cuda.is_available() else \"cpu\"\n",
    "\n",
    "test_res = []\n",
    "for i in range(100, 701, 100):\n",
    "    utils.setup_seed(42)\n",
    "    model = MultiSRAMP(halfseqlen=i).to(device)\n",
    "    loss_fn = TriLoss()\n",
    "    optimizer = torch.optim.Adam(model.parameters(), lr=1e-3)\n",
    "    scheduler = torch.optim.lr_scheduler.CosineAnnealingWarmRestarts(optimizer, 1, 2)\n",
    "\n",
    "    print(f'Start to train')\n",
    "    start = time.time()\n",
    "    train(model, loss_fn, optimizer, scheduler, traindl, testdl, epochs=31, device=device)\n",
    "    end = time.time()\n",
    "    test_res += [test_loop(testdl, model, device, loss_fn)]\n",
    "    test_res += [end-start]\n",
    "    print(end-start)\n",
    "    \n",
    "    torch.save(model, f'model/full_{i}_mature_for_time.model')\n",
    "    utils.save(test_res, f'model/full_{i}_mature_for_time.metrics')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "edc51cc7-3bf4-489b-a024-6c42faf72c8b",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_res = utils.load(f'model/full_700_mature_for_time.metrics')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "fd933df1-e264-4107-b67d-29415fba48fa",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 500x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "from matplotlib import font_manager\n",
    "\n",
    "font_manager.fontManager.addfont('/home/fanrui/fonts/ARIAL.TTF')\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "plt.rcParams['font.family'] = 'Arial'\n",
    "plt.rcParams['font.size'] = 10\n",
    "\n",
    "checkpoint = [i*2 for i in range(100, 701, 100)]\n",
    "aucs = [i[1] for i in test_res[::2]]\n",
    "ts = [i/31 for i in test_res[1::2]]\n",
    "\n",
    "fig = plt.figure(figsize=(5, 4))\n",
    "ax1 = plt.subplot()\n",
    "ax2 = ax1.twinx()\n",
    "ax1.plot(checkpoint, aucs, label='AUROC', color='#44757A')\n",
    "ax1.scatter(checkpoint, aucs, color='#44757A')\n",
    "ax2.plot(checkpoint, ts, color='#D44C3C', label='Time')\n",
    "ax2.scatter(checkpoint, ts, color='#D44C3C')\n",
    "\n",
    "ax1.set_ylabel('AUROC')\n",
    "ax2.set_ylabel('Training Time (seconds / epoch)')\n",
    "ax1.set_xlabel('Windows Size (nt)')\n",
    "ax1.set_ylim((0.935, 0.965))\n",
    "fig.legend(bbox_to_anchor=(.87, .3), frameon=False)\n",
    "plt.tight_layout()\n",
    "plt.savefig('fig/sfig1.svg')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8b5ddc40-6811-49e0-80cf-38a84c04fd03",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true
   },
   "source": [
    "## Threshold"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "5e797420-7dec-41da-ac3d-5f4c04b507c2",
   "metadata": {},
   "outputs": [],
   "source": [
    "traindf, testdf = utils.load('data/sramp1/sramp1_mature.data')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "5e307886-2f46-4f9d-8bee-010ddcccc43d",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 194716/194716 [01:44<00:00, 1867.09it/s]\n"
     ]
    }
   ],
   "source": [
    "testds = df2ds_multi(testdf)\n",
    "testdl = ds2dl(testds, batch_size=128, shuffle=False, num_workers=4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "44c6b6ff-73b8-4b7e-bd4a-3a774f6ead6a",
   "metadata": {},
   "outputs": [],
   "source": [
    "import time\n",
    "\n",
    "# print(f'Building model')\n",
    "device = \"cuda\" if torch.cuda.is_available() else \"cpu\"\n",
    "\n",
    "utils.setup_seed(42)\n",
    "model = MultiSRAMP(mode=i).to(device)\n",
    "model.load_state_dict(torch.load('model/full_400_mature.model'))\n",
    "loss_fn = TriLoss()\n",
    "optimizer = torch.optim.Adam(model.parameters(), lr=1e-3)\n",
    "scheduler = torch.optim.lr_scheduler.CosineAnnealingWarmRestarts(optimizer, 1, 2)\n",
    "\n",
    "print(f'Start to train')\n",
    "test_res = []\n",
    "test_res += [test_loop(testdl, model, device, loss_fn)]\n",
    "    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "3de047d0-f8cd-4be0-9d27-2cb88d50a356",
   "metadata": {},
   "outputs": [],
   "source": [
    "metric = utils.load('model/full_400_mature.metrics')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "0fdd2336-5874-4c5e-93a1-8e83c3fd2c71",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "target\t spec\t thrs\n",
      "0.5\t0.5013\t0.0000\n",
      "0.55\t0.5638\t0.0015\n",
      "0.6\t0.6052\t0.0079\n",
      "0.65\t0.6521\t0.0245\n",
      "0.7\t0.7003\t0.0655\n",
      "0.75\t0.7503\t0.1832\n",
      "0.8\t0.8002\t0.4042\n",
      "0.85\t0.8501\t0.6399\n",
      "0.9\t0.9000\t0.7927\n",
      "0.95\t0.9500\t0.8981\n"
     ]
    }
   ],
   "source": [
    "print('target\\t spec\\t thrs')\n",
    "for i in range(50, 100, 5):\n",
    "    i /= 100\n",
    "    spec = (1 - metric[0][3])\n",
    "    print(f'{i}\\t{spec[spec > i][-1]:.4f}\\t{metric[0][5][spec > i][-1]:.4f}')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "4beed8cb-d60f-474c-819c-d39f24dfd4a1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "inf\n",
      "1.0\n"
     ]
    }
   ],
   "source": [
    "for idx, i in enumerate(1 - metric[0][3]):\n",
    "    \n",
    "    if i > 0.8:\n",
    "        print(metric[0][5][idx])\n",
    "    if i > 0.9:\n",
    "        print(metric[0][5][idx])\n",
    "        break\n",
    "    print(i)\n",
    "    break"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "0d64a46e-5126-4762-885b-164e47c63ed1",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "7463"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(metric[0][5])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2bc109c2-0a41-4def-80b2-d75c297ccd5e",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true
   },
   "source": [
    "## Performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "87f92c20-56be-46d8-8427-b6389015fbaa",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 400x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 400x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 400x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 400x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "from matplotlib import font_manager\n",
    "\n",
    "font_manager.fontManager.addfont('/home/fanrui/fonts/ARIAL.TTF')\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "plt.rcParams['font.family'] = 'Arial'\n",
    "plt.rcParams['font.size'] = 10\n",
    "\n",
    "names = ['Genome Only', 'Sequence Only', 'deepSRAMP']\n",
    "\n",
    "\n",
    "for mode in ['mature', 'full']:\n",
    "    test_res = []\n",
    "    for i in ['genomeonly', 'seqonly', 'full']:\n",
    "        test_res += utils.load(f'model/{i}_400_{mode}.metrics')\n",
    "    \n",
    "    data = np.array([[i[-1]['1.0']['precision'], i[-1]['1.0']['recall'], i[-1]['1.0']['f1-score'], i[-1]['0.0']['recall'], i[-1]['accuracy'], i[-2], i[2], i[1]] for i in (test_res)])\n",
    "    t = pd.DataFrame(data, columns=['Precision', 'Recall', 'F1', 'Specificity', 'Accuracy', 'MCC', 'AUPRC', 'AUROC'], index=[i for i in ['genomeonly', 'seqonly', 'full']])\n",
    "    t.index.name = f'deepSRAMP[{mode}]'\n",
    "    t.to_csv(f'fig/fig1b_{mode}.csv')\n",
    "\n",
    "    # sramp1\n",
    "    if mode == 'mature':\n",
    "        a = pd.read_csv(f'sramp1_sramp1test.out', sep='\\t')\n",
    "    else:\n",
    "        a = pd.read_csv(f'sramp1_sramp1test_full.out', sep='\\t')\n",
    "    a = a[a.Position == 51]\n",
    "    a['y'] = a['Seq_ID'].astype(str).str.split('.', expand=True)[1].astype(int)\n",
    "    ratio = a.y.sum() / a.shape[0]\n",
    "    \n",
    "    # auc\n",
    "    fig = plt.figure(figsize=(4, 4))\n",
    "    ax = fig.subplots()\n",
    "    ax.spines['top'].set_visible(False)\n",
    "    ax.spines['right'].set_visible(False)\n",
    "    \n",
    "    tpr, fpr, ths = roc_curve(a.y, a['Score(Combined)'])\n",
    "    t1 = auc(tpr, fpr)\n",
    "    plt.plot(tpr, fpr, label=f'SRAMP AUROC={t1:.3f}', lw=2, color='#44757A')\n",
    "    \n",
    "    colors = ['#A5A58D', '#CB997E', '#D44C3C']\n",
    "    # colors = ['tab:orange', 'tab:purple', 'tab:red']\n",
    "    for idx, (name, r) in enumerate(zip(names, test_res)):\n",
    "        test_loss, aucc, auprcc, fpr, tpr, throc, prec, reca, thprc, mcc, report = r\n",
    "        plt.plot(fpr, tpr, label=f'{name} AUROC={aucc:.3f}', lw=2, color=colors[idx])\n",
    "    plt.legend(frameon=False)\n",
    "    plt.xlim(-0.01, 1.01)\n",
    "    plt.ylim(-0.01, 1.01)\n",
    "    plt.xlabel('False Postive Rate')\n",
    "    plt.ylabel('True Postive Rate')\n",
    "    plt.tight_layout()\n",
    "    plt.savefig(f'fig/fig1b_roc_{mode}.svg', transparent=True)\n",
    "    plt.show()\n",
    "    \n",
    "    # prc\n",
    "    fig = plt.figure(figsize=(4, 4))\n",
    "    ax = fig.subplots()\n",
    "    ax.spines['top'].set_visible(False)\n",
    "    ax.spines['right'].set_visible(False)\n",
    "    \n",
    "    tpr, fpr, ths = precision_recall_curve(a.y, a['Score(Combined)'])\n",
    "    t2 = auc(fpr, tpr)\n",
    "    # plt.plot(fpr, tpr, label=f'SRAMP1 AUC={t2:.3f}', lw=2)\n",
    "    plt.plot(fpr, tpr, label=f'SRAMP AUPRC={t2:.3f}', lw=2, color='#44757A')\n",
    "    \n",
    "    \n",
    "    for idx, (name, r) in enumerate(zip(names, test_res)):\n",
    "        test_loss, aucc, auprcc, fpr, tpr, throc, prec, reca, thprc, mcc, report = r\n",
    "        plt.plot(reca, prec, label=f'{name} AUPRC={auprcc:.3f}', lw=2, color=colors[idx])\n",
    "    plt.legend(loc=(0,0), frameon=False)\n",
    "    plt.xlim(-0.01, 1.01)\n",
    "    plt.ylim(-0.01, 1.01)\n",
    "    plt.xlabel('Recall')\n",
    "    plt.ylabel('Precision')\n",
    "    plt.tight_layout()\n",
    "    plt.savefig(f'fig/fig1b_prc_{mode}.svg', transparent=True)\n",
    "    plt.show()\n",
    "    # break\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0ef0c446-f552-4734-a61b-d30d4167d9fe",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "e0e2719a-b098-4bdd-b93c-83913011fbc0",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true
   },
   "source": [
    "## Ablation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "4a822db2-22db-4642-ba0c-1dd78210d04a",
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from matplotlib import font_manager\n",
    "\n",
    "font_manager.fontManager.addfont('/home/fanrui/fonts/ARIAL.TTF')\n",
    "plt.rcParams['svg.fonttype'] = 'none'\n",
    "plt.rcParams['font.family'] = 'Arial'\n",
    "plt.rcParams['font.size'] = 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "786ae2f7-bdba-4c58-ba9f-91b5c0e676e6",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_res = [utils.load(f'ablation/full_400_{i}.metrics')[0] for i in ['trans', 'gru']] + utils.load(f'model/full_400_mature.metrics')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "4e4fe742-3a34-4115-adfe-2b7fd1ab3012",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Precision</th>\n",
       "      <th>Recall</th>\n",
       "      <th>F1</th>\n",
       "      <th>Specificity</th>\n",
       "      <th>Accuracy</th>\n",
       "      <th>MCC</th>\n",
       "      <th>AUPRC</th>\n",
       "      <th>AUROC</th>\n",
       "      <th>method</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.338210</td>\n",
       "      <td>0.928479</td>\n",
       "      <td>0.495814</td>\n",
       "      <td>0.819210</td>\n",
       "      <td>0.829099</td>\n",
       "      <td>0.496423</td>\n",
       "      <td>0.605482</td>\n",
       "      <td>0.940652</td>\n",
       "      <td>w/o Transformer</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.102973</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.186718</td>\n",
       "      <td>0.133130</td>\n",
       "      <td>0.211586</td>\n",
       "      <td>0.117084</td>\n",
       "      <td>0.357254</td>\n",
       "      <td>0.842180</td>\n",
       "      <td>w/o BiGRU</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.339256</td>\n",
       "      <td>0.933488</td>\n",
       "      <td>0.497651</td>\n",
       "      <td>0.819081</td>\n",
       "      <td>0.829435</td>\n",
       "      <td>0.499280</td>\n",
       "      <td>0.624192</td>\n",
       "      <td>0.942970</td>\n",
       "      <td>Model</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Precision    Recall        F1  Specificity  Accuracy       MCC     AUPRC  \\\n",
       "0   0.338210  0.928479  0.495814     0.819210  0.829099  0.496423  0.605482   \n",
       "1   0.102973  1.000000  0.186718     0.133130  0.211586  0.117084  0.357254   \n",
       "2   0.339256  0.933488  0.497651     0.819081  0.829435  0.499280  0.624192   \n",
       "\n",
       "      AUROC           method  \n",
       "0  0.940652  w/o Transformer  \n",
       "1  0.842180        w/o BiGRU  \n",
       "2  0.942970            Model  "
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t = np.array([[i[-1]['1.0']['precision'], i[-1]['1.0']['recall'], i[-1]['1.0']['f1-score'], i[-1]['0.0']['recall'], i[-1]['accuracy'], i[-2], i[2], i[1]] for i in (test_res)])\n",
    "t = pd.DataFrame(t, columns=['Precision', 'Recall', 'F1', 'Specificity', 'Accuracy', 'MCC', 'AUPRC', 'AUROC'])\n",
    "method = ['w/o Transformer', 'w/o BiGRU', 'Model']\n",
    "t['method'] = method\n",
    "t"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "ff1da537-920a-48cd-9c70-8c6f7013c3ef",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 400x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(4, 4))\n",
    "ax = fig.subplots()\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "\n",
    "names = ['w/o Transformer', 'w/o BiGRU', 'deepSRAMP']\n",
    "colors = ['#44757A', '#CB997E', '#D44C3C'] # , '#A5A58D'\n",
    "for idx, (name, r) in enumerate(zip(names, test_res)):\n",
    "    test_loss, aucc, auprcc, fpr, tpr, throc, prec, reca, thprc, mcc, report = r = r\n",
    "    plt.plot(fpr, tpr, label=f'{name} AUROC={aucc:.3f}', lw=2, color=colors[idx])\n",
    "plt.legend(frameon=False)\n",
    "plt.xlim(-0.01, 1.01)\n",
    "plt.ylim(-0.01, 1.01)\n",
    "plt.xlabel('False Postive Rate')\n",
    "plt.ylabel('True Postive Rate')\n",
    "plt.tight_layout()\n",
    "plt.savefig(f'fig/sfig3_roc_ablation.svg', transparent=True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "d8e46661-6556-4ab6-a060-b8775d17adb4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 400x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(4, 4))\n",
    "ax = fig.subplots()\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "\n",
    "names = ['w/o Transformer', 'w/o BiGRU', 'deepSRAMP']\n",
    "colors = ['#44757A', '#CB997E', '#D44C3C'] # , '#A5A58D'\n",
    "for idx, (name, r) in enumerate(zip(names, test_res)):\n",
    "    test_loss, aucc, auprcc, fpr, tpr, throc, prec, reca, thprc, mcc, report = r = r\n",
    "    plt.plot(reca, prec, label=f'{name} AUPRC={auprcc:.3f}', lw=2, color=colors[idx])\n",
    "plt.legend(frameon=False)\n",
    "plt.xlim(-0.01, 1.01)\n",
    "plt.ylim(-0.01, 1.01)\n",
    "plt.xlabel('Recall')\n",
    "plt.ylabel('Precision')\n",
    "plt.tight_layout()\n",
    "plt.savefig(f'fig/sfig3_prc_ablation.svg', transparent=True)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3bd1905c-bd0d-493d-8734-702ac803cb2b",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "aaedcf61-488e-44b3-9e56-d1e639960f0b",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "ccad58d7-84a0-49d5-bd42-a5de703b9cb3",
   "metadata": {
    "jp-MarkdownHeadingCollapsed": true
   },
   "source": [
    "## Misc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "0b73f7bb-8ec9-497b-a871-0e2ac18b8e6d",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_res = [utils.load(f'ablation/full_400_{i}.metrics')[0] for i in ['trans', 'gru', 'gru2conv']] + utils.load(f'model/full_400_mature.metrics')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "07a303c4-4f58-4eff-a82b-a66d978a39a0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Precision</th>\n",
       "      <th>Recall</th>\n",
       "      <th>F1</th>\n",
       "      <th>Specificity</th>\n",
       "      <th>Accuracy</th>\n",
       "      <th>MCC</th>\n",
       "      <th>AUPRC</th>\n",
       "      <th>AUROC</th>\n",
       "      <th>method</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.338210</td>\n",
       "      <td>0.928479</td>\n",
       "      <td>0.495814</td>\n",
       "      <td>0.819210</td>\n",
       "      <td>0.829099</td>\n",
       "      <td>0.496423</td>\n",
       "      <td>0.605482</td>\n",
       "      <td>0.940652</td>\n",
       "      <td>w/o Transformer</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.102973</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.186718</td>\n",
       "      <td>0.133130</td>\n",
       "      <td>0.211586</td>\n",
       "      <td>0.117084</td>\n",
       "      <td>0.357254</td>\n",
       "      <td>0.842180</td>\n",
       "      <td>w/o GRU</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.102973</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.186718</td>\n",
       "      <td>0.133130</td>\n",
       "      <td>0.211586</td>\n",
       "      <td>0.117084</td>\n",
       "      <td>0.220883</td>\n",
       "      <td>0.778641</td>\n",
       "      <td>GRU -&gt; Conv</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.339256</td>\n",
       "      <td>0.933488</td>\n",
       "      <td>0.497651</td>\n",
       "      <td>0.819081</td>\n",
       "      <td>0.829435</td>\n",
       "      <td>0.499280</td>\n",
       "      <td>0.624192</td>\n",
       "      <td>0.942970</td>\n",
       "      <td>Model</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Precision    Recall        F1  Specificity  Accuracy       MCC     AUPRC  \\\n",
       "0   0.338210  0.928479  0.495814     0.819210  0.829099  0.496423  0.605482   \n",
       "1   0.102973  1.000000  0.186718     0.133130  0.211586  0.117084  0.357254   \n",
       "2   0.102973  1.000000  0.186718     0.133130  0.211586  0.117084  0.220883   \n",
       "3   0.339256  0.933488  0.497651     0.819081  0.829435  0.499280  0.624192   \n",
       "\n",
       "      AUROC           method  \n",
       "0  0.940652  w/o Transformer  \n",
       "1  0.842180          w/o GRU  \n",
       "2  0.778641      GRU -> Conv  \n",
       "3  0.942970            Model  "
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t = np.array([[i[-1]['1.0']['precision'], i[-1]['1.0']['recall'], i[-1]['1.0']['f1-score'], i[-1]['0.0']['recall'], i[-1]['accuracy'], i[-2], i[2], i[1]] for i in (test_res)])\n",
    "t = pd.DataFrame(t, columns=['Precision', 'Recall', 'F1', 'Specificity', 'Accuracy', 'MCC', 'AUPRC', 'AUROC'])\n",
    "method = ['w/o Transformer', 'w/o GRU', 'GRU -> Conv', 'Model']\n",
    "t['method'] = method\n",
    "t"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d9b061c7-1d2f-4683-ba46-10202a82eb19",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a8f55c03-59b9-4439-853b-20e189ee9947",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "ba6c54fe-7a4f-43dc-8cf3-a3ff2e927e16",
   "metadata": {},
   "outputs": [],
   "source": [
    "traindf, testdf = utils.load('data/sramp1/sramp1_mature.data')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "f337f2aa-e710-4b42-b703-ee64684be49b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6195121365921553"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(lambda x: x[x>1].shape[0] / x.shape[0])(traindf.grp.value_counts())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "4df8d98e-f2cf-4856-bf4f-5d0148227b90",
   "metadata": {},
   "outputs": [],
   "source": [
    "traindf, testdf = utils.load('data/sramp1/sramp1_full.data')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "4c2e751e-23e4-4c7f-90c1-d3a7c4029b85",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 297220/297220 [02:56<00:00, 1684.24it/s]\n"
     ]
    }
   ],
   "source": [
    "# trainds = df2ds_multi(traindf)\n",
    "# traindl = ds2dl(trainds, batch_size=128, drop_last=True, num_workers=4)\n",
    "\n",
    "testds = df2ds_multi(testdf)\n",
    "testdl = ds2dl(testds, batch_size=128, shuffle=False, num_workers=4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "bfa9d3a3-6040-4def-b3fa-c4710f2814f8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Start to train\n",
      "[Test] ACC: 76.6%, Loss: 0.498984, AUROC: 0.9015, AUPRC: 0.4630, Size: 119110, Ratio: 0.09050\n",
      "7.152557373046875e-07\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "\n",
    "# print(f'Building model')\n",
    "device = \"cuda\" if torch.cuda.is_available() else \"cpu\"\n",
    "\n",
    "test_res = []\n",
    "for i in ['genomeonly', 'seqonly', 'full'][-1:]:\n",
    "    utils.setup_seed(42)\n",
    "    model = MultiSRAMP(mode=i).to(device)\n",
    "    model.load_state_dict(torch.load('model/full_400_mature_rmute.model', weights_only=True))\n",
    "    loss_fn = TriLoss()\n",
    "    optimizer = torch.optim.Adam(model.parameters(), lr=1e-3)\n",
    "    scheduler = torch.optim.lr_scheduler.CosineAnnealingWarmRestarts(optimizer, 1, 2)\n",
    "\n",
    "    print(f'Start to train')\n",
    "    start = time.time()\n",
    "    # train(model, loss_fn, optimizer, scheduler, traindl, testdl, epochs=31, device=device)\n",
    "    end = time.time()\n",
    "    test_res += [test_loop(testdl, model, device, loss_fn)]\n",
    "    print(end-start)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "775cb2b5-d682-4ab7-8e1f-10b20f44f64d",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "9c8b1ee7-1053-4cca-8f98-e7b7038a27fe",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[(0.3712875876739765,\n",
       "  np.float64(0.942489751218833),\n",
       "  np.float64(0.6178487946059482),\n",
       "  array([0.        , 0.        , 0.        , ..., 0.99525524, 0.9952737 ,\n",
       "         1.        ]),\n",
       "  array([0.00000000e+00, 9.27643785e-05, 6.58627087e-03, ...,\n",
       "         1.00000000e+00, 1.00000000e+00, 1.00000000e+00]),\n",
       "  array([           inf, 9.98873770e-01, 9.98231351e-01, ...,\n",
       "         1.49521355e-08, 1.49409320e-08, 4.27142455e-09]),\n",
       "  array([0.09050458, 0.09050534, 0.0905061 , ..., 1.        , 1.        ,\n",
       "         1.        ]),\n",
       "  array([1.00000000e+00, 1.00000000e+00, 1.00000000e+00, ...,\n",
       "         1.85528757e-04, 9.27643785e-05, 0.00000000e+00]),\n",
       "  array([4.27142455e-09, 4.69060701e-09, 5.15317655e-09, ...,\n",
       "         9.98868227e-01, 9.98870194e-01, 9.98873770e-01]),\n",
       "  np.float64(0.4983852619852373),\n",
       "  {'0.0': {'precision': 0.9923944989118973,\n",
       "    'recall': 0.8166528200867719,\n",
       "    'f1-score': 0.8959873604893759,\n",
       "    'support': 108330.0},\n",
       "   '1.0': {'precision': 0.33713789881190764,\n",
       "    'recall': 0.9371057513914657,\n",
       "    'f1-score': 0.4958766935008836,\n",
       "    'support': 10780.0},\n",
       "   'accuracy': 0.8275543615145664,\n",
       "   'macro avg': {'precision': 0.6647661988619025,\n",
       "    'recall': 0.8768792857391188,\n",
       "    'f1-score': 0.6959320269951297,\n",
       "    'support': 119110.0},\n",
       "   'weighted avg': {'precision': 0.9330907784091865,\n",
       "    'recall': 0.8275543615145664,\n",
       "    'f1-score': 0.8597755143795954,\n",
       "    'support': 119110.0}})]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "utils.load(f'model/full_400_mature_rmute.metrics')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "73957453-f27d-4872-bd4a-c4ea757d856b",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "id": "1b6c3f15-5abb-4b4c-a274-2f27125b34f0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x400 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "tt = t.set_index(['method']).stack().reset_index()\n",
    "\n",
    "fig = plt.figure(figsize=(6, 4))\n",
    "ax = fig.subplots()\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "sns.barplot(data=tt, x='level_1', y=0, hue='method', palette=['#44757A', '#CB997E', '#D44C3C']) # , '#A5A58D'\n",
    "plt.legend(loc='upper center', frameon=False, ncol=2, bbox_to_anchor=(0.5, 1.2))\n",
    "plt.ylim((0, 0.7))\n",
    "plt.xlabel('')\n",
    "plt.ylabel('')\n",
    "plt.tight_layout()\n",
    "# plt.savefig('fig/fig2_allmetric.svg')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1311ecc8-21ec-49ec-8f89-0a2a81255cc1",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:sramp]",
   "language": "python",
   "name": "conda-env-sramp-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.12.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}