hierarchical_lda_wgcna.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Clustring other than hSBM\n",
    "Run LDA, wgcna, topic mapping and hierarchical"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Filippo_Valle \n",
      "last updated: Thu Nov 19 2020 \n",
      "\n",
      "CPython 3.8.6\n",
      "IPython 7.19.0\n",
      "\n",
      "pandas 1.1.4\n",
      "numpy 1.19.4\n",
      "seaborn 0.11.0\n",
      "tensorflow 2.3.1\n",
      "sklearn 0.0\n",
      "scipy 1.5.3\n",
      "topicpy 0.0.1\n",
      "\n",
      "compiler   : GCC 7.5.0\n",
      "system     : Linux\n",
      "release    : 5.4.39-linuxkit\n",
      "machine    : x86_64\n",
      "processor  : x86_64\n",
      "CPU cores  : 2\n",
      "interpreter: 64bit\n",
      "Git hash   : 9eaf7cf78429ca754e7b263692d8ab00cff33d2d\n",
      "Git repo   : git@github.com:fvalle1/cancers\n",
      "Git branch : develop_first_round\n",
      "watermark 2.0.2\n"
     ]
    }
   ],
   "source": [
    "%load_ext watermark\n",
    "%watermark -v -m  -u -n -p pandas,numpy,seaborn,tensorflow,sklearn,scipy,topicpy -a Filippo_Valle -g -r -b -w"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt  \n",
    "import pandas as pd  \n",
    "import numpy as np \n",
    "import seaborn as sns\n",
    "import os, sys\n",
    "from topicpy.hsbmpy import get_file, define_labels, get_cluster_given_l, get_max_available_L\n",
    "from topicpy.geneontology import topic_analysis\n",
    "import tensorflow as tf\n",
    "from sklearn.metrics import homogeneity_completeness_v_measure\n",
    "import scipy.cluster.hierarchy as shc\n",
    "from sklearn.cluster import AgglomerativeClustering\n",
    "from scipy.stats import pearsonr\n",
    "from sklearn.decomposition import LatentDirichletAllocation\n",
    "from topicpy.lda import lda"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "os.chdir('breast/tests/high_filters/')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'/home/jovyan/work/phd/cancers/breast/tests/high_filters'"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "level = 2\n",
    "directory=f\"{os.getcwd()}/\"\n",
    "#L=get_max_available_L(directory)\n",
    "os.chdir(directory)\n",
    "os.getcwd()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 3000 entries, ENSG00000268903 to ENSG00000251322\n",
      "Columns: 1000 entries, GTEX-111CU-0626-SM-5EGHL to GTEX-ZYY3-2226-SM-5E45A\n",
      "dtypes: float64(1000)\n",
      "memory usage: 22.9+ MB\n",
      "None\n",
      "Maximum expression value: 6.8e+05\n"
     ]
    },
    {
     "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>GTEX-111CU-0626-SM-5EGHL</th>\n",
       "      <th>GTEX-111CU-1226-SM-5EGIN</th>\n",
       "      <th>GTEX-111CU-1426-SM-5GZYP</th>\n",
       "      <th>GTEX-111VG-1626-SM-5EGIO</th>\n",
       "      <th>GTEX-111YS-2426-SM-5GZZQ</th>\n",
       "      <th>GTEX-1128S-0926-SM-5GZZU</th>\n",
       "      <th>GTEX-1128S-2126-SM-5H12U</th>\n",
       "      <th>GTEX-113IC-0426-SM-5HL5O</th>\n",
       "      <th>GTEX-117YW-0008-SM-5QGRD</th>\n",
       "      <th>GTEX-117YW-0326-SM-5N9CY</th>\n",
       "      <th>...</th>\n",
       "      <th>GTEX-ZVZQ-0006-SM-51MR8</th>\n",
       "      <th>GTEX-ZVZQ-0726-SM-51MR3</th>\n",
       "      <th>GTEX-ZWKS-0626-SM-5GIDV</th>\n",
       "      <th>GTEX-ZXG5-0926-SM-5NQ8H</th>\n",
       "      <th>GTEX-ZY6K-1326-SM-5GZWP</th>\n",
       "      <th>GTEX-ZYFG-0126-SM-5GIDH</th>\n",
       "      <th>GTEX-ZYFG-1826-SM-5GZWX</th>\n",
       "      <th>GTEX-ZYT6-0826-SM-5E43I</th>\n",
       "      <th>GTEX-ZYT6-2826-SM-5GICX</th>\n",
       "      <th>GTEX-ZYY3-2226-SM-5E45A</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>ENSG00000268903</th>\n",
       "      <td>1.807</td>\n",
       "      <td>2.518</td>\n",
       "      <td>3.326</td>\n",
       "      <td>2.670</td>\n",
       "      <td>3.002</td>\n",
       "      <td>8.406</td>\n",
       "      <td>10.11</td>\n",
       "      <td>0.2126</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.4232</td>\n",
       "      <td>...</td>\n",
       "      <td>50.82</td>\n",
       "      <td>1.303</td>\n",
       "      <td>7.593</td>\n",
       "      <td>2.683</td>\n",
       "      <td>2.141</td>\n",
       "      <td>1.660</td>\n",
       "      <td>1.481</td>\n",
       "      <td>1.317</td>\n",
       "      <td>1.459</td>\n",
       "      <td>10.97</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ENSG00000269981</th>\n",
       "      <td>3.478</td>\n",
       "      <td>5.765</td>\n",
       "      <td>7.898</td>\n",
       "      <td>5.162</td>\n",
       "      <td>7.022</td>\n",
       "      <td>11.630</td>\n",
       "      <td>14.18</td>\n",
       "      <td>0.1884</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.4018</td>\n",
       "      <td>...</td>\n",
       "      <td>48.22</td>\n",
       "      <td>1.668</td>\n",
       "      <td>11.070</td>\n",
       "      <td>1.560</td>\n",
       "      <td>2.846</td>\n",
       "      <td>1.876</td>\n",
       "      <td>1.895</td>\n",
       "      <td>2.706</td>\n",
       "      <td>1.837</td>\n",
       "      <td>10.79</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2 rows × 1000 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                 GTEX-111CU-0626-SM-5EGHL  GTEX-111CU-1226-SM-5EGIN  \\\n",
       "ENSG00000268903                     1.807                     2.518   \n",
       "ENSG00000269981                     3.478                     5.765   \n",
       "\n",
       "                 GTEX-111CU-1426-SM-5GZYP  GTEX-111VG-1626-SM-5EGIO  \\\n",
       "ENSG00000268903                     3.326                     2.670   \n",
       "ENSG00000269981                     7.898                     5.162   \n",
       "\n",
       "                 GTEX-111YS-2426-SM-5GZZQ  GTEX-1128S-0926-SM-5GZZU  \\\n",
       "ENSG00000268903                     3.002                     8.406   \n",
       "ENSG00000269981                     7.022                    11.630   \n",
       "\n",
       "                 GTEX-1128S-2126-SM-5H12U  GTEX-113IC-0426-SM-5HL5O  \\\n",
       "ENSG00000268903                     10.11                    0.2126   \n",
       "ENSG00000269981                     14.18                    0.1884   \n",
       "\n",
       "                 GTEX-117YW-0008-SM-5QGRD  GTEX-117YW-0326-SM-5N9CY  ...  \\\n",
       "ENSG00000268903                       0.0                    0.4232  ...   \n",
       "ENSG00000269981                       0.0                    0.4018  ...   \n",
       "\n",
       "                 GTEX-ZVZQ-0006-SM-51MR8  GTEX-ZVZQ-0726-SM-51MR3  \\\n",
       "ENSG00000268903                    50.82                    1.303   \n",
       "ENSG00000269981                    48.22                    1.668   \n",
       "\n",
       "                 GTEX-ZWKS-0626-SM-5GIDV  GTEX-ZXG5-0926-SM-5NQ8H  \\\n",
       "ENSG00000268903                    7.593                    2.683   \n",
       "ENSG00000269981                   11.070                    1.560   \n",
       "\n",
       "                 GTEX-ZY6K-1326-SM-5GZWP  GTEX-ZYFG-0126-SM-5GIDH  \\\n",
       "ENSG00000268903                    2.141                    1.660   \n",
       "ENSG00000269981                    2.846                    1.876   \n",
       "\n",
       "                 GTEX-ZYFG-1826-SM-5GZWX  GTEX-ZYT6-0826-SM-5E43I  \\\n",
       "ENSG00000268903                    1.481                    1.317   \n",
       "ENSG00000269981                    1.895                    2.706   \n",
       "\n",
       "                 GTEX-ZYT6-2826-SM-5GICX  GTEX-ZYY3-2226-SM-5E45A  \n",
       "ENSG00000268903                    1.459                    10.97  \n",
       "ENSG00000269981                    1.837                    10.79  \n",
       "\n",
       "[2 rows x 1000 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.read_csv(\"mainTable.csv\", index_col=[0], header=[0]).dropna()\n",
    "totalobjcets = len(df.columns)\n",
    "print(df.info())\n",
    "print(\"Maximum expression value: %.1e\"%df.max().max())\n",
    "df.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "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>SMATSSCR</th>\n",
       "      <th>SMCENTER</th>\n",
       "      <th>SMPTHNTS</th>\n",
       "      <th>SMRIN</th>\n",
       "      <th>SMTS</th>\n",
       "      <th>SMTSD</th>\n",
       "      <th>SMUBRID</th>\n",
       "      <th>SMTSISCH</th>\n",
       "      <th>SMTSPAX</th>\n",
       "      <th>SMNABTCH</th>\n",
       "      <th>...</th>\n",
       "      <th>SME1ANTI</th>\n",
       "      <th>SMSPLTRD</th>\n",
       "      <th>SMBSMMRT</th>\n",
       "      <th>SME1SNSE</th>\n",
       "      <th>SME1PCTS</th>\n",
       "      <th>SMRRNART</th>\n",
       "      <th>SME1MPRT</th>\n",
       "      <th>SMNUM5CD</th>\n",
       "      <th>SMDPMPRT</th>\n",
       "      <th>SME2PCTS</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SAMPID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>GTEX-1117F-0003-SM-58Q7G</th>\n",
       "      <td>NaN</td>\n",
       "      <td>B1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Blood</td>\n",
       "      <td>Whole Blood</td>\n",
       "      <td>0013756</td>\n",
       "      <td>1188.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>BP-38516</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>GTEX-1117F-0003-SM-5DWSB</th>\n",
       "      <td>NaN</td>\n",
       "      <td>B1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Blood</td>\n",
       "      <td>Whole Blood</td>\n",
       "      <td>0013756</td>\n",
       "      <td>1188.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>BP-38516</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>GTEX-1117F-0003-SM-6WBT7</th>\n",
       "      <td>NaN</td>\n",
       "      <td>B1</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Blood</td>\n",
       "      <td>Whole Blood</td>\n",
       "      <td>0013756</td>\n",
       "      <td>1188.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>BP-38516</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>3 rows × 62 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                          SMATSSCR SMCENTER SMPTHNTS  SMRIN   SMTS  \\\n",
       "SAMPID                                                               \n",
       "GTEX-1117F-0003-SM-58Q7G       NaN       B1      NaN    NaN  Blood   \n",
       "GTEX-1117F-0003-SM-5DWSB       NaN       B1      NaN    NaN  Blood   \n",
       "GTEX-1117F-0003-SM-6WBT7       NaN       B1      NaN    NaN  Blood   \n",
       "\n",
       "                                SMTSD  SMUBRID  SMTSISCH  SMTSPAX  SMNABTCH  \\\n",
       "SAMPID                                                                        \n",
       "GTEX-1117F-0003-SM-58Q7G  Whole Blood  0013756    1188.0      NaN  BP-38516   \n",
       "GTEX-1117F-0003-SM-5DWSB  Whole Blood  0013756    1188.0      NaN  BP-38516   \n",
       "GTEX-1117F-0003-SM-6WBT7  Whole Blood  0013756    1188.0      NaN  BP-38516   \n",
       "\n",
       "                          ... SME1ANTI SMSPLTRD SMBSMMRT SME1SNSE SME1PCTS  \\\n",
       "SAMPID                    ...                                                \n",
       "GTEX-1117F-0003-SM-58Q7G  ...      NaN      NaN      NaN      NaN      NaN   \n",
       "GTEX-1117F-0003-SM-5DWSB  ...      NaN      NaN      NaN      NaN      NaN   \n",
       "GTEX-1117F-0003-SM-6WBT7  ...      NaN      NaN      NaN      NaN      NaN   \n",
       "\n",
       "                         SMRRNART SME1MPRT  SMNUM5CD  SMDPMPRT  SME2PCTS  \n",
       "SAMPID                                                                    \n",
       "GTEX-1117F-0003-SM-58Q7G      NaN      NaN       NaN       NaN       NaN  \n",
       "GTEX-1117F-0003-SM-5DWSB      NaN      NaN       NaN       NaN       NaN  \n",
       "GTEX-1117F-0003-SM-6WBT7      NaN      NaN       NaN       NaN       NaN  \n",
       "\n",
       "[3 rows x 62 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_files = pd.read_csv(\"files.dat\", index_col=[0])\n",
    "df_files.head(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "true_out = []\n",
    "label = \"SMTS\"\n",
    "if label not in df_files.columns:\n",
    "    raise AttributeError(f\"{label} not Avaliable\")\n",
    "for sample in df.columns.values:\n",
    "    try:\n",
    "        true_out.append(get_file(sample, df_files)[label])\n",
    "    except:\n",
    "        print(*sys.exc_info())\n",
    "        true_out.append('unknown')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "scores = {}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([  1,   8,  29, 978])"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "try:\n",
    "    with open(\"clustersizes.txt\",'r') as f:\n",
    "        xl = np.sort(np.array(f.read().split(sep='\\n'))[:-1].astype(int))\n",
    "except:\n",
    "        xl=np.linspace(2,50,3, dtype=int)\n",
    "xl"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 1,  8, 29])"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "xl=xl[:-1]\n",
    "xl"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# hierarchical"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig=plt.figure()\n",
    "dend = shc.dendrogram(shc.linkage(np.log2(df.T.values+1), method='average'), leaf_rotation=90., leaf_font_size=8.,)\n",
    "plt.xlabel(\"samples\", fontsize=16)\n",
    "plt.show()\n",
    "fig.savefig(\"hierarchical_dendogram.pdf\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "hierarchical-log\n",
      "testing with 1 clusters\n",
      "saving clusters\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.7/site-packages/ipykernel_launcher.py:24: DeprecationWarning: Support for multi-dimensional indexing (e.g. `index[:, None]`) on an Index is deprecated and will be removed in a future version.  Convert to a numpy array before indexing instead.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "testing with 8 clusters\n",
      "saving clusters\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.7/site-packages/ipykernel_launcher.py:24: DeprecationWarning: Support for multi-dimensional indexing (e.g. `index[:, None]`) on an Index is deprecated and will be removed in a future version.  Convert to a numpy array before indexing instead.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "testing with 29 clusters\n",
      "saving clusters\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.7/site-packages/ipykernel_launcher.py:24: DeprecationWarning: Support for multi-dimensional indexing (e.g. `index[:, None]`) on an Index is deprecated and will be removed in a future version.  Convert to a numpy array before indexing instead.\n"
     ]
    }
   ],
   "source": [
    "#hierarchical\n",
    "scores['hierarchical']={\n",
    "    'h':[],\n",
    "    'c':[],\n",
    "    'V':[]\n",
    "}\n",
    "def pearson_affinity(M, parallel=True):\n",
    "    return 1. - np.array([[pearsonr(a,b)[0] for a in M] for b in M])\n",
    "\n",
    "print(\"hierarchical-log\")\n",
    "os.system('mkdir -p hierarchical-log')\n",
    "hierarchical_model = AgglomerativeClustering(n_clusters=1, affinity='euclidean', linkage='complete')  \n",
    "for l,x in enumerate(xl):\n",
    "    print(\"testing with %d clusters\"%x)\n",
    "    hierarchical_model.n_clusters=x\n",
    "    data = np.log2(1.+df.T.values)\n",
    "    #data = df.T.values\n",
    "    out = hierarchical_model.fit_predict(data)\n",
    "        \n",
    "    #save clusters\n",
    "    print(\"saving clusters\")\n",
    "    df_clusters = pd.DataFrame(index=np.arange(len(df.columns)))\n",
    "    for c in np.arange(out.max()+1)[::-1]:\n",
    "        c_objects = df.columns[np.argwhere(out==c)].T[0]\n",
    "        df_clusters.insert(0,\"Cluster %d\"%(c+1),np.concatenate((c_objects,[np.nan for _ in np.arange(len(df.columns)-len(c_objects))])))\n",
    "    df_clusters.dropna(axis=0,how='all', inplace=True)\n",
    "    df_clusters.to_csv(\"hierarchical-log/hierarchical-log_level_%d_clusters.csv\"%(l), index=False, header=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# LDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(\"topicsizes.txt\",'r') as f:\n",
    "    tl = np.array(f.read().split(sep='\\n'))[:-1].astype(int)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 1  8 29]\n"
     ]
    }
   ],
   "source": [
    "xl = xl[:3]\n",
    "print(xl)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": true,
    "jupyter": {
     "outputs_hidden": true
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "model created\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 3000 entries, ENSG00000268903 to ENSG00000251322\n",
      "Columns: 1000 entries, GTEX-111CU-0626-SM-5EGHL to GTEX-ZYY3-2226-SM-5E45A\n",
      "dtypes: float64(1000)\n",
      "memory usage: 22.9+ MB\n",
      "None\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 22951 entries, GTEX-1117F-0003-SM-58Q7G to K-562-SM-E9EZZ\n",
      "Data columns (total 62 columns):\n",
      " #   Column     Non-Null Count  Dtype  \n",
      "---  ------     --------------  -----  \n",
      " 0   SMATSSCR   15739 non-null  float64\n",
      " 1   SMCENTER   22724 non-null  object \n",
      " 2   SMPTHNTS   15737 non-null  object \n",
      " 3   SMRIN      19998 non-null  float64\n",
      " 4   SMTS       22951 non-null  object \n",
      " 5   SMTSD      22951 non-null  object \n",
      " 6   SMUBRID    22951 non-null  object \n",
      " 7   SMTSISCH   22650 non-null  float64\n",
      " 8   SMTSPAX    15715 non-null  float64\n",
      " 9   SMNABTCH   22575 non-null  object \n",
      " 10  SMNABTCHT  22575 non-null  object \n",
      " 11  SMNABTCHD  22575 non-null  object \n",
      " 12  SMGEBTCH   22228 non-null  object \n",
      " 13  SMGEBTCHD  22209 non-null  object \n",
      " 14  SMGEBTCHT  22951 non-null  object \n",
      " 15  SMAFRZE    22349 non-null  object \n",
      " 16  SMGTC      379 non-null    object \n",
      " 17  SME2MPRT   18859 non-null  float64\n",
      " 18  SMCHMPRS   18859 non-null  float64\n",
      " 19  SMNTRART   18859 non-null  float64\n",
      " 20  SMNUMGPS   0 non-null      float64\n",
      " 21  SMMAPRT    18859 non-null  float64\n",
      " 22  SMEXNCRT   18859 non-null  float64\n",
      " 23  SM550NRM   0 non-null      float64\n",
      " 24  SMGNSDTC   18859 non-null  float64\n",
      " 25  SMUNMPRT   18859 non-null  float64\n",
      " 26  SM350NRM   0 non-null      float64\n",
      " 27  SMRDLGTH   18859 non-null  float64\n",
      " 28  SMMNCPB    0 non-null      float64\n",
      " 29  SME1MMRT   18859 non-null  float64\n",
      " 30  SMSFLGTH   18859 non-null  float64\n",
      " 31  SMESTLBS   18859 non-null  float64\n",
      " 32  SMMPPD     18859 non-null  float64\n",
      " 33  SMNTERRT   18859 non-null  float64\n",
      " 34  SMRRNANM   18859 non-null  float64\n",
      " 35  SMRDTTL    18859 non-null  float64\n",
      " 36  SMVQCFL    18859 non-null  float64\n",
      " 37  SMMNCV     0 non-null      float64\n",
      " 38  SMTRSCPT   18859 non-null  float64\n",
      " 39  SMMPPDPR   18859 non-null  float64\n",
      " 40  SMCGLGTH   0 non-null      float64\n",
      " 41  SMGAPPCT   0 non-null      float64\n",
      " 42  SMUNPDRD   18859 non-null  float64\n",
      " 43  SMNTRNRT   18859 non-null  float64\n",
      " 44  SMMPUNRT   18859 non-null  float64\n",
      " 45  SMEXPEFF   18859 non-null  float64\n",
      " 46  SMMPPDUN   18859 non-null  float64\n",
      " 47  SME2MMRT   18859 non-null  float64\n",
      " 48  SME2ANTI   18859 non-null  float64\n",
      " 49  SMALTALG   18859 non-null  float64\n",
      " 50  SME2SNSE   18859 non-null  float64\n",
      " 51  SMMFLGTH   18859 non-null  float64\n",
      " 52  SME1ANTI   18859 non-null  float64\n",
      " 53  SMSPLTRD   18859 non-null  float64\n",
      " 54  SMBSMMRT   18859 non-null  float64\n",
      " 55  SME1SNSE   18859 non-null  float64\n",
      " 56  SME1PCTS   18859 non-null  float64\n",
      " 57  SMRRNART   18859 non-null  float64\n",
      " 58  SME1MPRT   18859 non-null  float64\n",
      " 59  SMNUM5CD   0 non-null      float64\n",
      " 60  SMDPMPRT   18859 non-null  float64\n",
      " 61  SME2PCTS   18859 non-null  float64\n",
      "dtypes: float64(49), object(13)\n",
      "memory usage: 11.0+ MB\n",
      "None\n",
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 22951 entries, GTEX-1117F-0003-SM-58Q7G to K-562-SM-E9EZZ\n",
      "Data columns (total 62 columns):\n",
      " #   Column     Non-Null Count  Dtype  \n",
      "---  ------     --------------  -----  \n",
      " 0   SMATSSCR   15739 non-null  float64\n",
      " 1   SMCENTER   22724 non-null  object \n",
      " 2   SMPTHNTS   15737 non-null  object \n",
      " 3   SMRIN      19998 non-null  float64\n",
      " 4   SMTS       22951 non-null  object \n",
      " 5   SMTSD      22951 non-null  object \n",
      " 6   SMUBRID    22951 non-null  object \n",
      " 7   SMTSISCH   22650 non-null  float64\n",
      " 8   SMTSPAX    15715 non-null  float64\n",
      " 9   SMNABTCH   22575 non-null  object \n",
      " 10  SMNABTCHT  22575 non-null  object \n",
      " 11  SMNABTCHD  22575 non-null  object \n",
      " 12  SMGEBTCH   22228 non-null  object \n",
      " 13  SMGEBTCHD  22209 non-null  object \n",
      " 14  SMGEBTCHT  22951 non-null  object \n",
      " 15  SMAFRZE    22349 non-null  object \n",
      " 16  SMGTC      379 non-null    object \n",
      " 17  SME2MPRT   18859 non-null  float64\n",
      " 18  SMCHMPRS   18859 non-null  float64\n",
      " 19  SMNTRART   18859 non-null  float64\n",
      " 20  SMNUMGPS   0 non-null      float64\n",
      " 21  SMMAPRT    18859 non-null  float64\n",
      " 22  SMEXNCRT   18859 non-null  float64\n",
      " 23  SM550NRM   0 non-null      float64\n",
      " 24  SMGNSDTC   18859 non-null  float64\n",
      " 25  SMUNMPRT   18859 non-null  float64\n",
      " 26  SM350NRM   0 non-null      float64\n",
      " 27  SMRDLGTH   18859 non-null  float64\n",
      " 28  SMMNCPB    0 non-null      float64\n",
      " 29  SME1MMRT   18859 non-null  float64\n",
      " 30  SMSFLGTH   18859 non-null  float64\n",
      " 31  SMESTLBS   18859 non-null  float64\n",
      " 32  SMMPPD     18859 non-null  float64\n",
      " 33  SMNTERRT   18859 non-null  float64\n",
      " 34  SMRRNANM   18859 non-null  float64\n",
      " 35  SMRDTTL    18859 non-null  float64\n",
      " 36  SMVQCFL    18859 non-null  float64\n",
      " 37  SMMNCV     0 non-null      float64\n",
      " 38  SMTRSCPT   18859 non-null  float64\n",
      " 39  SMMPPDPR   18859 non-null  float64\n",
      " 40  SMCGLGTH   0 non-null      float64\n",
      " 41  SMGAPPCT   0 non-null      float64\n",
      " 42  SMUNPDRD   18859 non-null  float64\n",
      " 43  SMNTRNRT   18859 non-null  float64\n",
      " 44  SMMPUNRT   18859 non-null  float64\n",
      " 45  SMEXPEFF   18859 non-null  float64\n",
      " 46  SMMPPDUN   18859 non-null  float64\n",
      " 47  SME2MMRT   18859 non-null  float64\n",
      " 48  SME2ANTI   18859 non-null  float64\n",
      " 49  SMALTALG   18859 non-null  float64\n",
      " 50  SME2SNSE   18859 non-null  float64\n",
      " 51  SMMFLGTH   18859 non-null  float64\n",
      " 52  SME1ANTI   18859 non-null  float64\n",
      " 53  SMSPLTRD   18859 non-null  float64\n",
      " 54  SMBSMMRT   18859 non-null  float64\n",
      " 55  SME1SNSE   18859 non-null  float64\n",
      " 56  SME1PCTS   18859 non-null  float64\n",
      " 57  SMRRNART   18859 non-null  float64\n",
      " 58  SME1MPRT   18859 non-null  float64\n",
      " 59  SMNUM5CD   0 non-null      float64\n",
      " 60  SMDPMPRT   18859 non-null  float64\n",
      " 61  SME2PCTS   18859 non-null  float64\n",
      "dtypes: float64(49), object(13)\n",
      "memory usage: 11.7+ MB\n",
      "None\n",
      "lda\n",
      "testing with 1 clusters and 1 topics\n",
      "lda()\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    6.8s remaining:    6.8s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    6.9s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.0s remaining:    0.0s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.0s remaining:    0.0s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.0s remaining:    0.0s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iteration: 1 of max_iter: 5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.0s remaining:    0.0s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.0s remaining:    0.0s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.0s remaining:    0.0s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iteration: 2 of max_iter: 5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.2s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.2s remaining:    0.2s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.2s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.2s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.2s remaining:    0.2s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.2s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iteration: 3 of max_iter: 5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.2s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.0s remaining:    0.0s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.0s remaining:    0.0s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iteration: 4 of max_iter: 5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.2s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.2s remaining:    0.2s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.2s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.2s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.2s remaining:    0.2s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.2s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.0s remaining:    0.0s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iteration: 5 of max_iter: 5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.8s remaining:    0.8s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.9s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.3s remaining:    0.3s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.4s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "saving word-distr\n",
      "saving topic-distr\n",
      "saving clusters\n",
      "saving metrics\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.2s remaining:    0.2s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.2s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "testing with 8 clusters and 8 topics\n",
      "lda()\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iteration: 1 of max_iter: 5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.1s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.2s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.2s remaining:    0.2s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.2s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.3s remaining:    0.3s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.4s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.5s remaining:    0.5s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.6s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.2s remaining:    0.2s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iteration: 2 of max_iter: 5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.2s remaining:    0.2s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.3s remaining:    0.3s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.2s remaining:    0.2s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.2s remaining:    0.2s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.2s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.3s remaining:    0.3s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.2s remaining:    0.2s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.2s remaining:    0.2s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.2s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.2s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iteration: 3 of max_iter: 5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.2s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.2s remaining:    0.2s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.2s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.2s remaining:    0.2s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.2s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.1s remaining:    0.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.2s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.3s remaining:    0.3s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.4s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.2s remaining:    0.2s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.4s remaining:    0.4s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.4s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.8s remaining:    0.8s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.8s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iteration: 4 of max_iter: 5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.7s remaining:    0.7s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.9s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.4s remaining:    0.4s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.4s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.8s remaining:    0.8s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.8s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.3s remaining:    0.3s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.5s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.3s remaining:    0.3s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.2s remaining:    0.2s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.4s remaining:    0.4s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.6s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.3s remaining:    0.3s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iteration: 5 of max_iter: 5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    1.3s remaining:    1.3s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    1.4s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.3s remaining:    0.3s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.4s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "saving word-distr\n",
      "saving topic-distr\n",
      "saving clusters\n",
      "saving metrics\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    1.1s remaining:    1.1s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    1.2s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "testing with 29 clusters and 29 topics\n",
      "lda()\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.4s remaining:    0.4s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.5s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.3s remaining:    0.3s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.7s remaining:    0.7s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.8s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.5s remaining:    0.5s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.6s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.6s remaining:    0.6s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.7s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.5s remaining:    0.5s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.8s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.5s remaining:    0.5s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.5s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.2s remaining:    0.2s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iteration: 1 of max_iter: 5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.5s remaining:    0.5s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.5s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.5s remaining:    0.5s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.5s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.3s remaining:    0.3s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.3s remaining:    0.3s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.3s remaining:    0.3s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.2s remaining:    0.2s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.2s remaining:    0.2s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.2s remaining:    0.2s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iteration: 2 of max_iter: 5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.3s remaining:    0.3s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.3s remaining:    0.3s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.3s remaining:    0.3s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.4s remaining:    0.4s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.5s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.4s remaining:    0.4s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.6s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.5s remaining:    0.5s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.6s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.5s remaining:    0.5s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.5s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.4s remaining:    0.4s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.4s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iteration: 3 of max_iter: 5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.4s remaining:    0.4s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.4s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.5s remaining:    0.5s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.5s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.4s remaining:    0.4s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.4s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.5s remaining:    0.5s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.6s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.5s remaining:    0.5s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.5s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.3s remaining:    0.3s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.4s remaining:    0.4s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.4s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.4s remaining:    0.4s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.5s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iteration: 4 of max_iter: 5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.4s remaining:    0.4s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.5s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.4s remaining:    0.4s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.5s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.3s remaining:    0.3s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.5s remaining:    0.5s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.6s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    1.0s remaining:    1.0s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    1.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    1.6s remaining:    1.6s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    1.7s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.7s remaining:    0.7s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.8s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    0.7s remaining:    0.7s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    0.8s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "iteration: 5 of max_iter: 5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    2.0s remaining:    2.0s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    2.3s finished\n",
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    1.9s remaining:    1.9s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    2.5s finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "saving word-distr\n",
      "saving topic-distr\n",
      "saving clusters\n",
      "saving metrics\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=4)]: Using backend LokyBackend with 4 concurrent workers.\n",
      "[Parallel(n_jobs=4)]: Done   2 out of   4 | elapsed:    1.9s remaining:    1.9s\n",
      "[Parallel(n_jobs=4)]: Done   4 out of   4 | elapsed:    2.1s finished\n"
     ]
    }
   ],
   "source": [
    "model=lda(n_jobs=4, verbose=2,  max_iter=5)\n",
    "model.full_analysis(directory, xl, tl=None, logarithmise=False, label=\"SMTS\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# WGCNA"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Use [WGCNA.ipynb](WGCNA.ipynb) to run analyses before continuing and store data in the next step"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "saving clusters\n",
      "saving clusters\n",
      "saving clusters\n",
      "saving clusters\n"
     ]
    }
   ],
   "source": [
    "from topicpy.hsbmpy import get_max_available_L\n",
    "for l in range(get_max_available_L(directory, \"wgcna\")+1):\n",
    "    df_wgcna = pd.read_csv(\"wgcna/wgcna_level_%d_labels.csv\"%l, index_col=0)\n",
    "    totalobjcets = len(df_wgcna.index)\n",
    "    out = df_wgcna['x'].values\n",
    "    print(\"saving clusters\")\n",
    "    df_clusters = pd.DataFrame(index=np.arange(totalobjcets))\n",
    "    for c in np.arange(out.max()+1)[::-1]:\n",
    "        c_objects = df_wgcna.index.values[np.argwhere(out==(c+1))].T[0]\n",
    "        df_clusters.insert(0,\"Cluster %d\"%(c+1),np.concatenate((c_objects,[np.nan for _ in np.arange(totalobjcets-len(c_objects))])))\n",
    "    df_clusters.dropna(axis=0,how='all', inplace=True)\n",
    "    df_clusters.to_csv(\"wgcna/wgcna_level_%d_clusters.csv\"%l, index=False, header=True)\n",
    "df_wgcna_td = pd.read_csv(\"wgcna/wgcna_level_0_topic-dist.csv\")\n",
    "df_wgcna_td.columns.values[0]='doc'\n",
    "df_wgcna_td.index.name='i_doc'\n",
    "df_wgcna_td.to_csv(\"wgcna/wgcna_level_0_topic-dist.csv\", index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_word_distr = pd.read_csv(\"wgcna/wgcna_level_0_word-dist.csv\", index_col=0)\n",
    "df_topics = pd.DataFrame()\n",
    "max_L = df_word_distr.shape[0]\n",
    "for topic in df_word_distr.columns[::-1]:\n",
    "    t_series = df_word_distr[topic]\n",
    "    t_series = t_series[t_series>0.5]\n",
    "    df_topics.insert(0,topic.replace(\"MM\",\"ME\"),np.concatenate((t_series.index.values,np.repeat(np.nan, df_word_distr.shape[0]-len(t_series)))))\n",
    "df_topics.dropna(how=\"all\", axis=0).to_csv(\"wgcna/wgcna_level_0_topics.csv\", index=False, header=True)"
   ]
  },
  {
   "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": "3.8.6"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {},
    "version_major": 2,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}