std-scalar-100mb-.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Scale Features and Build Model\n",
    "### Scales Raw Features\n",
    "\n",
    "Import CSV of Aggregated Darshan Logs <br>\n",
    "Apply Log10 and Percent Scaling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import math\n",
    "import matplotlib.pyplot as plt\n",
    "import random\n",
    "\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"./raws.csv\",lineterminator='\\n',sep = ',' ,error_bad_lines=False)\n",
    "#df.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.drop(df.columns[0],axis = 1)\n",
    "df = df.drop(df.columns[0],axis = 1)\n",
    "f = pd.DataFrame()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "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>posix_read_time</th>\n",
       "      <th>posix_write_time</th>\n",
       "      <th>posix_meta_time</th>\n",
       "      <th>posix_bytes_read</th>\n",
       "      <th>posix_bytes_read_100</th>\n",
       "      <th>posix_bytes_read_1K</th>\n",
       "      <th>posix_bytes_read_10K</th>\n",
       "      <th>posix_bytes_read_100K</th>\n",
       "      <th>posix_bytes_read_1lM</th>\n",
       "      <th>posix_bytes_read_4M</th>\n",
       "      <th>...</th>\n",
       "      <th>posix_m_align</th>\n",
       "      <th>lustre_number_of_files</th>\n",
       "      <th>lustre_mdts</th>\n",
       "      <th>lustre_osts</th>\n",
       "      <th>lustre_stripe_size</th>\n",
       "      <th>lustre_stripe_offset</th>\n",
       "      <th>lustre_stripe_width</th>\n",
       "      <th>lustre_number_of_osts</th>\n",
       "      <th>jobid</th>\n",
       "      <th>path</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>36.000000</td>\n",
       "      <td>1048576.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>9042391.0</td>\n",
       "      <td>../results/USER375904U_sss2_id9042391_9-26-414...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>104.611641</td>\n",
       "      <td>10.024055</td>\n",
       "      <td>20.060841</td>\n",
       "      <td>2.390891e+10</td>\n",
       "      <td>147688.0</td>\n",
       "      <td>141853.0</td>\n",
       "      <td>33215.0</td>\n",
       "      <td>261267.0</td>\n",
       "      <td>51049.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>13704.0</td>\n",
       "      <td>176.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>360.000000</td>\n",
       "      <td>1048576.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>176.0</td>\n",
       "      <td>147.0</td>\n",
       "      <td>9868242.0</td>\n",
       "      <td>../results/USER382862U_SpEC_id9868242_4-29-741...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>124.560730</td>\n",
       "      <td>42.051125</td>\n",
       "      <td>54.839272</td>\n",
       "      <td>5.019637e+10</td>\n",
       "      <td>332059.0</td>\n",
       "      <td>684893.0</td>\n",
       "      <td>130143.0</td>\n",
       "      <td>919688.0</td>\n",
       "      <td>2515.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>6704.0</td>\n",
       "      <td>224.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>360.000000</td>\n",
       "      <td>1048576.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>224.0</td>\n",
       "      <td>186.0</td>\n",
       "      <td>9029683.0</td>\n",
       "      <td>../results/USER264205U_SpEC_id9029683_9-20-580...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>25763.292969</td>\n",
       "      <td>582.297363</td>\n",
       "      <td>24.895737</td>\n",
       "      <td>5.488943e+12</td>\n",
       "      <td>30785.0</td>\n",
       "      <td>767.0</td>\n",
       "      <td>508381.0</td>\n",
       "      <td>2040.0</td>\n",
       "      <td>13683.0</td>\n",
       "      <td>626216.0</td>\n",
       "      <td>...</td>\n",
       "      <td>2320.0</td>\n",
       "      <td>290.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>54.993103</td>\n",
       "      <td>1048576.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>290.0</td>\n",
       "      <td>52.0</td>\n",
       "      <td>9165166.0</td>\n",
       "      <td>../results/USER341996U_CCTM_v52_Linux2_x86_64i...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>154.534821</td>\n",
       "      <td>681.548279</td>\n",
       "      <td>658.484985</td>\n",
       "      <td>2.293203e+10</td>\n",
       "      <td>588029.0</td>\n",
       "      <td>2926599.0</td>\n",
       "      <td>425127.0</td>\n",
       "      <td>336133.0</td>\n",
       "      <td>98.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>6328.0</td>\n",
       "      <td>201.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>360.000000</td>\n",
       "      <td>1048576.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>201.0</td>\n",
       "      <td>158.0</td>\n",
       "      <td>8269019.0</td>\n",
       "      <td>../results/USER141481U_SpEC_id8269019_3-3-4542...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>875282</th>\n",
       "      <td>138.354477</td>\n",
       "      <td>82.278084</td>\n",
       "      <td>194.485565</td>\n",
       "      <td>5.593977e+10</td>\n",
       "      <td>216146.0</td>\n",
       "      <td>535152.0</td>\n",
       "      <td>99814.0</td>\n",
       "      <td>889010.0</td>\n",
       "      <td>12485.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>9920.0</td>\n",
       "      <td>124.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>360.000000</td>\n",
       "      <td>1048576.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>124.0</td>\n",
       "      <td>110.0</td>\n",
       "      <td>7759072.0</td>\n",
       "      <td>../results/USER336298U_SpEC_id7759072_10-30-62...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>875283</th>\n",
       "      <td>54.443073</td>\n",
       "      <td>231.440857</td>\n",
       "      <td>25.271391</td>\n",
       "      <td>1.465277e+09</td>\n",
       "      <td>3099.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>180662.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>8704.0</td>\n",
       "      <td>1088.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>360.000000</td>\n",
       "      <td>1048576.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1088.0</td>\n",
       "      <td>359.0</td>\n",
       "      <td>7275819.0</td>\n",
       "      <td>../results/USER331454U_pw.x_id7275819_8-2-4790...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>875284</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>36.000000</td>\n",
       "      <td>1048576.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>8305062.0</td>\n",
       "      <td>../results/USER371284U_sss2_id8305062_3-14-211...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>875285</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>8.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>36.000000</td>\n",
       "      <td>1048576.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>31428.0</td>\n",
       "      <td>../results/USER303986U_testnodes_opt.x_id31428...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>875286</th>\n",
       "      <td>227.063828</td>\n",
       "      <td>191.747269</td>\n",
       "      <td>172.671997</td>\n",
       "      <td>1.077412e+11</td>\n",
       "      <td>775359.0</td>\n",
       "      <td>1545229.0</td>\n",
       "      <td>305757.0</td>\n",
       "      <td>1897233.0</td>\n",
       "      <td>1068.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>...</td>\n",
       "      <td>15104.0</td>\n",
       "      <td>128.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>360.000000</td>\n",
       "      <td>1048576.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>128.0</td>\n",
       "      <td>118.0</td>\n",
       "      <td>7525550.0</td>\n",
       "      <td>../results/USER382862U_SpEC_id7525550_9-24-380...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>875287 rows × 50 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        posix_read_time  posix_write_time  posix_meta_time  posix_bytes_read  \\\n",
       "0              0.000000          0.000000         0.000000      0.000000e+00   \n",
       "1            104.611641         10.024055        20.060841      2.390891e+10   \n",
       "2            124.560730         42.051125        54.839272      5.019637e+10   \n",
       "3          25763.292969        582.297363        24.895737      5.488943e+12   \n",
       "4            154.534821        681.548279       658.484985      2.293203e+10   \n",
       "...                 ...               ...              ...               ...   \n",
       "875282       138.354477         82.278084       194.485565      5.593977e+10   \n",
       "875283        54.443073        231.440857        25.271391      1.465277e+09   \n",
       "875284         0.000000          0.000000         0.000000      0.000000e+00   \n",
       "875285         0.000000          0.000000         0.000000      0.000000e+00   \n",
       "875286       227.063828        191.747269       172.671997      1.077412e+11   \n",
       "\n",
       "        posix_bytes_read_100  posix_bytes_read_1K  posix_bytes_read_10K  \\\n",
       "0                        0.0                  0.0                   0.0   \n",
       "1                   147688.0             141853.0               33215.0   \n",
       "2                   332059.0             684893.0              130143.0   \n",
       "3                    30785.0                767.0              508381.0   \n",
       "4                   588029.0            2926599.0              425127.0   \n",
       "...                      ...                  ...                   ...   \n",
       "875282              216146.0             535152.0               99814.0   \n",
       "875283                3099.0                  0.0              180662.0   \n",
       "875284                   0.0                  0.0                   0.0   \n",
       "875285                   0.0                  0.0                   0.0   \n",
       "875286              775359.0            1545229.0              305757.0   \n",
       "\n",
       "        posix_bytes_read_100K  posix_bytes_read_1lM  posix_bytes_read_4M  ...  \\\n",
       "0                         0.0                   0.0                  0.0  ...   \n",
       "1                    261267.0               51049.0                  0.0  ...   \n",
       "2                    919688.0                2515.0                  0.0  ...   \n",
       "3                      2040.0               13683.0             626216.0  ...   \n",
       "4                    336133.0                  98.0                  0.0  ...   \n",
       "...                       ...                   ...                  ...  ...   \n",
       "875282               889010.0               12485.0                  0.0  ...   \n",
       "875283                    0.0                   0.0                  0.0  ...   \n",
       "875284                    0.0                   0.0                  0.0  ...   \n",
       "875285                    0.0                   0.0                  0.0  ...   \n",
       "875286              1897233.0                1068.0                  0.0  ...   \n",
       "\n",
       "        posix_m_align  lustre_number_of_files  lustre_mdts  lustre_osts  \\\n",
       "0                 0.0                     4.0          1.0    36.000000   \n",
       "1             13704.0                   176.0          1.0   360.000000   \n",
       "2              6704.0                   224.0          1.0   360.000000   \n",
       "3              2320.0                   290.0          1.0    54.993103   \n",
       "4              6328.0                   201.0          1.0   360.000000   \n",
       "...               ...                     ...          ...          ...   \n",
       "875282         9920.0                   124.0          1.0   360.000000   \n",
       "875283         8704.0                  1088.0          1.0   360.000000   \n",
       "875284            0.0                     5.0          1.0    36.000000   \n",
       "875285            8.0                     2.0          1.0    36.000000   \n",
       "875286        15104.0                   128.0          1.0   360.000000   \n",
       "\n",
       "        lustre_stripe_size  lustre_stripe_offset  lustre_stripe_width  \\\n",
       "0                1048576.0                   0.0                  4.0   \n",
       "1                1048576.0                   0.0                176.0   \n",
       "2                1048576.0                   0.0                224.0   \n",
       "3                1048576.0                   0.0                290.0   \n",
       "4                1048576.0                   0.0                201.0   \n",
       "...                    ...                   ...                  ...   \n",
       "875282           1048576.0                   0.0                124.0   \n",
       "875283           1048576.0                   0.0               1088.0   \n",
       "875284           1048576.0                   0.0                  5.0   \n",
       "875285           1048576.0                   0.0                  2.0   \n",
       "875286           1048576.0                   0.0                128.0   \n",
       "\n",
       "        lustre_number_of_osts      jobid  \\\n",
       "0                         4.0  9042391.0   \n",
       "1                       147.0  9868242.0   \n",
       "2                       186.0  9029683.0   \n",
       "3                        52.0  9165166.0   \n",
       "4                       158.0  8269019.0   \n",
       "...                       ...        ...   \n",
       "875282                  110.0  7759072.0   \n",
       "875283                  359.0  7275819.0   \n",
       "875284                    5.0  8305062.0   \n",
       "875285                    1.0    31428.0   \n",
       "875286                  118.0  7525550.0   \n",
       "\n",
       "                                                     path  \n",
       "0       ../results/USER375904U_sss2_id9042391_9-26-414...  \n",
       "1       ../results/USER382862U_SpEC_id9868242_4-29-741...  \n",
       "2       ../results/USER264205U_SpEC_id9029683_9-20-580...  \n",
       "3       ../results/USER341996U_CCTM_v52_Linux2_x86_64i...  \n",
       "4       ../results/USER141481U_SpEC_id8269019_3-3-4542...  \n",
       "...                                                   ...  \n",
       "875282  ../results/USER336298U_SpEC_id7759072_10-30-62...  \n",
       "875283  ../results/USER331454U_pw.x_id7275819_8-2-4790...  \n",
       "875284  ../results/USER371284U_sss2_id8305062_3-14-211...  \n",
       "875285  ../results/USER303986U_testnodes_opt.x_id31428...  \n",
       "875286  ../results/USER382862U_SpEC_id7525550_9-24-380...  \n",
       "\n",
       "[875287 rows x 50 columns]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['posix_read_time', 'posix_write_time', 'posix_meta_time',\n",
       "       'posix_bytes_read', 'posix_bytes_read_100', 'posix_bytes_read_1K',\n",
       "       'posix_bytes_read_10K', 'posix_bytes_read_100K', 'posix_bytes_read_1lM',\n",
       "       'posix_bytes_read_4M', 'posix_bytes_read_10M', 'posix_bytes_read_100M',\n",
       "       'posix_bytes_read_1G', 'posix_bytes_read_PLUS', 'posix_bytes_write',\n",
       "       'posix_bytes_write_100', 'posix_bytes_write_1K',\n",
       "       'posix_bytes_write_10K', 'posix_bytes_write_100K',\n",
       "       'posix_bytes_write_1M', 'posix_bytes_write_4M', 'posix_bytes_write_10M',\n",
       "       'posix_bytes_write_100M', 'posix_bytes_write_1G',\n",
       "       'posix_bytes_write_PLUS', 'posix_opens', 'posix_reads', 'posix_writes',\n",
       "       'posix_seeks', 'posix_stats', 'posix_mmaps', 'posix_fsyncs',\n",
       "       'posix_fdsyncs', 'posix_rename_sources', 'posix_rename_targets',\n",
       "       'posix_renamed_from', 'posix_renamed_mode', 'posix_number_of_files',\n",
       "       'nprocs', 'posix_f_align', 'posix_m_align', 'lustre_number_of_files',\n",
       "       'lustre_mdts', 'lustre_osts', 'lustre_stripe_size',\n",
       "       'lustre_stripe_offset', 'lustre_stripe_width', 'lustre_number_of_osts',\n",
       "       'jobid', 'path'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "df = df.dropna(axis=0, how='any')\n",
    "df.columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "#files\n",
    "f['log10_p_files'] = df['posix_number_of_files'] \n",
    "f['log10_l_files'] = df['lustre_number_of_files']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0               0.0\n",
       "1          880136.0\n",
       "2         2379598.0\n",
       "3         8903411.0\n",
       "4         7846387.0\n",
       "            ...    \n",
       "875282    2234152.0\n",
       "875283     197651.0\n",
       "875284          0.0\n",
       "875285          0.0\n",
       "875286    6065006.0\n",
       "Name: log10_p_accesses, Length: 875287, dtype: float64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#accesses\n",
    "df['p_accesses'] = df['posix_reads'] + df['posix_writes']\n",
    "f['log10_p_accesses'] = df['p_accesses']\n",
    "\n",
    "f['log10_p_accesses']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "#bytes\n",
    "f['log10_p_bytes'] = df['posix_bytes_read'] "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "f['log10_p_opens'] = df['posix_opens']\n",
    "f['log10_p_seeks'] = df['posix_seeks']\n",
    "f['log10_p_stats'] = df['posix_stats']\n",
    "f['log10_p_mode'] = df['posix_renamed_mode']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "f['log10_l_n_osts'] = df['lustre_number_of_osts']\n",
    "f['log10_l_stripe_w'] = df['lustre_stripe_width']\n",
    "f['log10_l_mdts'] = df['lustre_mdts']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "f['log10_p_nprocs'] = df['nprocs']\n",
    "f['log10_p_falign'] = df['posix_f_align']\n",
    "f['log10_p_malign'] = df['posix_m_align']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "f['perc_p_reads'] = df['posix_reads'] \n",
    "f['perc_p_writes'] = df['posix_writes'] "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "f['perc_p_bytes_read_100'] = df['posix_bytes_read_100']\n",
    "f['perc_p_bytes_read_1K'] = df['posix_bytes_read_1K']\n",
    "f['perc_p_bytes_read_10K'] = df['posix_bytes_read_10K'] \n",
    "f['perc_p_bytes_read_100K'] = df['posix_bytes_read_100K']\n",
    "f['perc_p_bytes_read_1M'] = df['posix_bytes_read_1lM']\n",
    "f['perc_p_bytes_read_4M'] = df['posix_bytes_read_4M']\n",
    "f['perc_p_bytes_read_10M'] = df['posix_bytes_read_10M']\n",
    "f['perc_p_bytes_read_100M'] = df['posix_bytes_read_100M']\n",
    "f['perc_p_bytes_read_1G'] = df['posix_bytes_read_1G']\n",
    "f['perc_p_bytes_read_PLUS'] = df['posix_bytes_read_PLUS']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "f['perc_p_bytes_write_100'] = df['posix_bytes_write_100']\n",
    "f['perc_p_bytes_write_1K'] = df['posix_bytes_write_1K']\n",
    "f['perc_p_bytes_write_10K'] = df['posix_bytes_write_10K']\n",
    "f['perc_p_bytes_write_100K'] = df['posix_bytes_write_100K']\n",
    "f['perc_p_bytes_write_1M'] = df['posix_bytes_write_1M']\n",
    "f['perc_p_bytes_write_4M'] = df['posix_bytes_write_4M']\n",
    "f['perc_p_bytes_write_10M'] = df['posix_bytes_write_10M']\n",
    "f['perc_p_bytes_write_100M'] = df['posix_bytes_write_100M']\n",
    "f['perc_p_bytes_write_1G'] = df['posix_bytes_write_1G']\n",
    "f['perc_p_bytes_write_PLUS'] = df['posix_bytes_write_PLUS']\n",
    "\n",
    "\n",
    "f = f.replace(-np.inf, -1)\n",
    "f = f.replace(np.nan, 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['time'] = df['posix_write_time'].astype('float') + df['posix_read_time'].astype('float') + df['posix_meta_time'].astype('float')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "df['bytes'] = df['posix_bytes_read'].astype('float') + df['posix_bytes_write'].astype('float')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df[df['bytes'] >99999999]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "f['throughput'] = df['bytes'].astype('float') / df['time']\n",
    "\n",
    "f = f[f['throughput'] >0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "log10_p_files              1.219270e+05\n",
       "log10_l_files              1.219270e+05\n",
       "log10_p_accesses           2.251942e+10\n",
       "log10_p_bytes              3.038456e+14\n",
       "log10_p_opens              5.531094e+08\n",
       "log10_p_seeks              1.445220e+10\n",
       "log10_p_stats              6.522921e+07\n",
       "log10_p_mode               5.337293e+07\n",
       "log10_l_n_osts             3.600000e+02\n",
       "log10_l_stripe_w           7.438575e+06\n",
       "log10_l_mdts               1.000000e+00\n",
       "log10_p_nprocs             3.200000e+05\n",
       "log10_p_falign             1.422540e+11\n",
       "log10_p_malign             1.085312e+06\n",
       "perc_p_reads               2.237846e+10\n",
       "perc_p_writes              1.302770e+10\n",
       "perc_p_bytes_read_100      5.221517e+08\n",
       "perc_p_bytes_read_1K       2.074657e+10\n",
       "perc_p_bytes_read_10K      1.536278e+09\n",
       "perc_p_bytes_read_100K     1.515506e+08\n",
       "perc_p_bytes_read_1M       4.044503e+08\n",
       "perc_p_bytes_read_4M       6.561462e+07\n",
       "perc_p_bytes_read_10M      2.083200e+06\n",
       "perc_p_bytes_read_100M     2.872090e+05\n",
       "perc_p_bytes_read_1G       1.792000e+06\n",
       "perc_p_bytes_write_100     1.302770e+10\n",
       "perc_p_bytes_write_1K      2.852127e+09\n",
       "perc_p_bytes_write_10K     3.867477e+08\n",
       "perc_p_bytes_write_100K    8.347452e+07\n",
       "perc_p_bytes_write_1M      1.357245e+07\n",
       "perc_p_bytes_write_4M      3.839488e+06\n",
       "perc_p_bytes_write_10M     6.190660e+05\n",
       "perc_p_bytes_write_100M    1.249280e+06\n",
       "perc_p_bytes_write_1G      1.937500e+04\n",
       "throughput                 2.344536e+09\n",
       "dtype: float64"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#delete columns with all zeros\n",
    "f = f.loc[:, (f != 0).any(axis=0)]\n",
    "\n",
    "#remove infinite values\n",
    "f = f.replace([np.inf, -np.inf], np.nan).dropna(axis=0)\n",
    "\n",
    "f.max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "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>log10_p_files</th>\n",
       "      <th>log10_l_files</th>\n",
       "      <th>log10_p_accesses</th>\n",
       "      <th>log10_p_bytes</th>\n",
       "      <th>log10_p_opens</th>\n",
       "      <th>log10_p_seeks</th>\n",
       "      <th>log10_p_stats</th>\n",
       "      <th>log10_p_mode</th>\n",
       "      <th>log10_l_n_osts</th>\n",
       "      <th>log10_l_stripe_w</th>\n",
       "      <th>...</th>\n",
       "      <th>perc_p_bytes_read_1G</th>\n",
       "      <th>perc_p_bytes_write_100</th>\n",
       "      <th>perc_p_bytes_write_1K</th>\n",
       "      <th>perc_p_bytes_write_10K</th>\n",
       "      <th>perc_p_bytes_write_100K</th>\n",
       "      <th>perc_p_bytes_write_1M</th>\n",
       "      <th>perc_p_bytes_write_4M</th>\n",
       "      <th>perc_p_bytes_write_10M</th>\n",
       "      <th>perc_p_bytes_write_100M</th>\n",
       "      <th>perc_p_bytes_write_1G</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>799.0</td>\n",
       "      <td>176.0</td>\n",
       "      <td>880136.0</td>\n",
       "      <td>2.390891e+10</td>\n",
       "      <td>8858.0</td>\n",
       "      <td>319241.0</td>\n",
       "      <td>34901.0</td>\n",
       "      <td>37230.0</td>\n",
       "      <td>147.0</td>\n",
       "      <td>176.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>202780.0</td>\n",
       "      <td>16078.0</td>\n",
       "      <td>21989.0</td>\n",
       "      <td>3156.0</td>\n",
       "      <td>1061.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>360.0</td>\n",
       "      <td>224.0</td>\n",
       "      <td>2379598.0</td>\n",
       "      <td>5.019637e+10</td>\n",
       "      <td>62398.0</td>\n",
       "      <td>1107764.0</td>\n",
       "      <td>270222.0</td>\n",
       "      <td>27594.0</td>\n",
       "      <td>186.0</td>\n",
       "      <td>224.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>18852.0</td>\n",
       "      <td>264906.0</td>\n",
       "      <td>22246.0</td>\n",
       "      <td>4282.0</td>\n",
       "      <td>14.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>290.0</td>\n",
       "      <td>290.0</td>\n",
       "      <td>8903411.0</td>\n",
       "      <td>5.488943e+12</td>\n",
       "      <td>8711.0</td>\n",
       "      <td>2010273.0</td>\n",
       "      <td>28432.0</td>\n",
       "      <td>115194.0</td>\n",
       "      <td>52.0</td>\n",
       "      <td>290.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>7302089.0</td>\n",
       "      <td>52045.0</td>\n",
       "      <td>73.0</td>\n",
       "      <td>9.0</td>\n",
       "      <td>25830.0</td>\n",
       "      <td>756.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>319.0</td>\n",
       "      <td>201.0</td>\n",
       "      <td>7846387.0</td>\n",
       "      <td>2.293203e+10</td>\n",
       "      <td>23158.0</td>\n",
       "      <td>6015926.0</td>\n",
       "      <td>400399.0</td>\n",
       "      <td>31536.0</td>\n",
       "      <td>158.0</td>\n",
       "      <td>201.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>489873.0</td>\n",
       "      <td>2821931.0</td>\n",
       "      <td>240372.0</td>\n",
       "      <td>17983.0</td>\n",
       "      <td>242.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>428.0</td>\n",
       "      <td>190.0</td>\n",
       "      <td>6647935.0</td>\n",
       "      <td>5.209185e+10</td>\n",
       "      <td>69261.0</td>\n",
       "      <td>4608438.0</td>\n",
       "      <td>410846.0</td>\n",
       "      <td>38982.0</td>\n",
       "      <td>158.0</td>\n",
       "      <td>190.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>255023.0</td>\n",
       "      <td>1927630.0</td>\n",
       "      <td>186048.0</td>\n",
       "      <td>11509.0</td>\n",
       "      <td>158.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>875280</th>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>57344.0</td>\n",
       "      <td>3.006477e+10</td>\n",
       "      <td>1808.0</td>\n",
       "      <td>59128.0</td>\n",
       "      <td>12.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28672.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>875281</th>\n",
       "      <td>13.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>102439.0</td>\n",
       "      <td>7.467916e+08</td>\n",
       "      <td>193.0</td>\n",
       "      <td>91.0</td>\n",
       "      <td>385.0</td>\n",
       "      <td>4818.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>4.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>33.0</td>\n",
       "      <td>23.0</td>\n",
       "      <td>10927.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>875282</th>\n",
       "      <td>624.0</td>\n",
       "      <td>124.0</td>\n",
       "      <td>2234152.0</td>\n",
       "      <td>5.593977e+10</td>\n",
       "      <td>35112.0</td>\n",
       "      <td>1035457.0</td>\n",
       "      <td>159180.0</td>\n",
       "      <td>16644.0</td>\n",
       "      <td>110.0</td>\n",
       "      <td>124.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>51807.0</td>\n",
       "      <td>397643.0</td>\n",
       "      <td>27690.0</td>\n",
       "      <td>3198.0</td>\n",
       "      <td>1207.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>875283</th>\n",
       "      <td>1088.0</td>\n",
       "      <td>1088.0</td>\n",
       "      <td>197651.0</td>\n",
       "      <td>1.465277e+09</td>\n",
       "      <td>2112.0</td>\n",
       "      <td>12509.0</td>\n",
       "      <td>4198.0</td>\n",
       "      <td>473916.0</td>\n",
       "      <td>359.0</td>\n",
       "      <td>1088.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>28.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>38.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>13824.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>875286</th>\n",
       "      <td>582.0</td>\n",
       "      <td>128.0</td>\n",
       "      <td>6065006.0</td>\n",
       "      <td>1.077412e+11</td>\n",
       "      <td>119432.0</td>\n",
       "      <td>2750807.0</td>\n",
       "      <td>578508.0</td>\n",
       "      <td>17520.0</td>\n",
       "      <td>118.0</td>\n",
       "      <td>128.0</td>\n",
       "      <td>...</td>\n",
       "      <td>0.0</td>\n",
       "      <td>216962.0</td>\n",
       "      <td>1239000.0</td>\n",
       "      <td>77141.0</td>\n",
       "      <td>6985.0</td>\n",
       "      <td>272.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>588408 rows × 34 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        log10_p_files  log10_l_files  log10_p_accesses  log10_p_bytes  \\\n",
       "1               799.0          176.0          880136.0   2.390891e+10   \n",
       "2               360.0          224.0         2379598.0   5.019637e+10   \n",
       "3               290.0          290.0         8903411.0   5.488943e+12   \n",
       "4               319.0          201.0         7846387.0   2.293203e+10   \n",
       "6               428.0          190.0         6647935.0   5.209185e+10   \n",
       "...               ...            ...               ...            ...   \n",
       "875280            1.0            1.0           57344.0   3.006477e+10   \n",
       "875281           13.0            4.0          102439.0   7.467916e+08   \n",
       "875282          624.0          124.0         2234152.0   5.593977e+10   \n",
       "875283         1088.0         1088.0          197651.0   1.465277e+09   \n",
       "875286          582.0          128.0         6065006.0   1.077412e+11   \n",
       "\n",
       "        log10_p_opens  log10_p_seeks  log10_p_stats  log10_p_mode  \\\n",
       "1              8858.0       319241.0        34901.0       37230.0   \n",
       "2             62398.0      1107764.0       270222.0       27594.0   \n",
       "3              8711.0      2010273.0        28432.0      115194.0   \n",
       "4             23158.0      6015926.0       400399.0       31536.0   \n",
       "6             69261.0      4608438.0       410846.0       38982.0   \n",
       "...               ...            ...            ...           ...   \n",
       "875280         1808.0        59128.0           12.0           0.0   \n",
       "875281          193.0           91.0          385.0        4818.0   \n",
       "875282        35112.0      1035457.0       159180.0       16644.0   \n",
       "875283         2112.0        12509.0         4198.0      473916.0   \n",
       "875286       119432.0      2750807.0       578508.0       17520.0   \n",
       "\n",
       "        log10_l_n_osts  log10_l_stripe_w  ...  perc_p_bytes_read_1G  \\\n",
       "1                147.0             176.0  ...                   0.0   \n",
       "2                186.0             224.0  ...                   0.0   \n",
       "3                 52.0             290.0  ...                   0.0   \n",
       "4                158.0             201.0  ...                   0.0   \n",
       "6                158.0             190.0  ...                   0.0   \n",
       "...                ...               ...  ...                   ...   \n",
       "875280             1.0               8.0  ...                   0.0   \n",
       "875281             4.0               4.0  ...                   0.0   \n",
       "875282           110.0             124.0  ...                   0.0   \n",
       "875283           359.0            1088.0  ...                   0.0   \n",
       "875286           118.0             128.0  ...                   0.0   \n",
       "\n",
       "        perc_p_bytes_write_100  perc_p_bytes_write_1K  perc_p_bytes_write_10K  \\\n",
       "1                     202780.0                16078.0                 21989.0   \n",
       "2                      18852.0               264906.0                 22246.0   \n",
       "3                    7302089.0                52045.0                    73.0   \n",
       "4                     489873.0              2821931.0                240372.0   \n",
       "6                     255023.0              1927630.0                186048.0   \n",
       "...                        ...                    ...                     ...   \n",
       "875280                     0.0                    0.0                     0.0   \n",
       "875281                    33.0                   23.0                 10927.0   \n",
       "875282                 51807.0               397643.0                 27690.0   \n",
       "875283                    28.0                    0.0                    38.0   \n",
       "875286                216962.0              1239000.0                 77141.0   \n",
       "\n",
       "        perc_p_bytes_write_100K  perc_p_bytes_write_1M  perc_p_bytes_write_4M  \\\n",
       "1                        3156.0                 1061.0                    0.0   \n",
       "2                        4282.0                   14.0                    0.0   \n",
       "3                           9.0                25830.0                  756.0   \n",
       "4                       17983.0                  242.0                    0.0   \n",
       "6                       11509.0                  158.0                    0.0   \n",
       "...                         ...                    ...                    ...   \n",
       "875280                      0.0                28672.0                    0.0   \n",
       "875281                      0.0                    0.0                    0.0   \n",
       "875282                   3198.0                 1207.0                    0.0   \n",
       "875283                      0.0                13824.0                    0.0   \n",
       "875286                   6985.0                  272.0                    0.0   \n",
       "\n",
       "        perc_p_bytes_write_10M  perc_p_bytes_write_100M  perc_p_bytes_write_1G  \n",
       "1                          0.0                      0.0                    0.0  \n",
       "2                          0.0                      0.0                    0.0  \n",
       "3                          0.0                      0.0                    0.0  \n",
       "4                          0.0                      0.0                    0.0  \n",
       "6                          0.0                      0.0                    0.0  \n",
       "...                        ...                      ...                    ...  \n",
       "875280                     0.0                      0.0                    0.0  \n",
       "875281                     0.0                      0.0                    0.0  \n",
       "875282                     0.0                      0.0                    0.0  \n",
       "875283                     0.0                      0.0                    0.0  \n",
       "875286                     0.0                      0.0                    0.0  \n",
       "\n",
       "[588408 rows x 34 columns]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t = pd.DataFrame()\n",
    "t['throughput'] = f['throughput']\n",
    "f = f.drop(labels = 'throughput', axis = 1)\n",
    "f"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df[df.index.isin(t.index)]\n",
    "t = t.reset_index()\n",
    "f = f.reset_index()\n",
    "f = f.drop(f.columns[0] , axis =1)\n",
    "t = t.drop(t.columns[0] , axis =1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "f = StandardScaler().fit_transform(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "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>throughput</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1.803194e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.282342e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.083669e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.724841e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4.581690e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>588403</th>\n",
       "      <td>5.482767e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>588404</th>\n",
       "      <td>2.136637e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>588405</th>\n",
       "      <td>1.362498e+08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>588406</th>\n",
       "      <td>1.942413e+07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>588407</th>\n",
       "      <td>1.842942e+08</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>588408 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "          throughput\n",
       "0       1.803194e+08\n",
       "1       2.282342e+08\n",
       "2       2.083669e+08\n",
       "3       1.724841e+07\n",
       "4       4.581690e+07\n",
       "...              ...\n",
       "588403  5.482767e+08\n",
       "588404  2.136637e+07\n",
       "588405  1.362498e+08\n",
       "588406  1.942413e+07\n",
       "588407  1.842942e+08\n",
       "\n",
       "[588408 rows x 1 columns]"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "t"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "throughput    76.744419\n",
      "dtype: float64\n",
      "throughput    2.344536e+09\n",
      "dtype: float64\n"
     ]
    }
   ],
   "source": [
    "print(t.min())\n",
    "print(t.max())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "rseed = 0\n",
    "t_size = 0.1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "fig = plt.figure()\n",
    "fig.suptitle('Bytes Read vs. Bytes Written Shaded by Throughput, Total Bytes Transferred >= 100MB', fontsize=14, fontweight='bold')\n",
    "\n",
    "ax = fig.add_subplot(111)\n",
    "sp = ax.scatter(df['posix_bytes_read'],df['posix_bytes_write'], marker = 'x',c = np.log10(t['throughput']),cmap='viridis')\n",
    "\n",
    "ax.set_xlabel('Bytes Read s')\n",
    "ax.set_ylabel('Bytes Written s')\n",
    "ax.loglog()\n",
    "#plt.autoscale(enable=True, axis='y')\n",
    "plt.xlim(10**0,10**15)\n",
    "plt.ylim(10**0,10**15)\n",
    "fig.colorbar(sp)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import r2_score\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from sklearn.metrics import mean_absolute_error\n",
    "from sklearn.metrics import mean_squared_log_error\n",
    "from sklearn.metrics import mean_absolute_percentage_error\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2304\n",
      "Mean True Value: \t 95557030\n",
      "Mean Absolute Error: \t 57193142\n",
      "Mean Squared Error:  9159394993309680.0\n",
      "Root Mean Squared Error:  95704728.16590454\n",
      "MAPE :13.90732356627179\n",
      "R2: 0.2762398672198274\n",
      "\n",
      "4858\n",
      "Mean True Value: \t 95211860\n",
      "Mean Absolute Error: \t 57447149\n",
      "Mean Squared Error:  9696730441393262.0\n",
      "Root Mean Squared Error:  98471977.9500405\n",
      "MAPE :19.971745753429353\n",
      "R2: 0.2340923727738773\n",
      "\n",
      "1932\n",
      "Mean True Value: \t 94940374\n",
      "Mean Absolute Error: \t 57047431\n",
      "Mean Squared Error:  9058964564376018.0\n",
      "Root Mean Squared Error:  95178592.99430738\n",
      "MAPE :13.643806571266154\n",
      "R2: 0.2717333610312159\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "for i in range(3):\n",
    "    \n",
    "    rseed = random.randint(1,10000)\n",
    "    print(rseed)\n",
    "    train_data, test_data, train_labels, test_labels = train_test_split(f,t, test_size=t_size, random_state=rseed)\n",
    "\n",
    "    reg = LinearRegression().fit(train_data, train_labels)\n",
    "\n",
    "    predicted_labels = reg.predict(test_data)\n",
    "\n",
    "    print(\"Mean True Value: \\t\",int(test_labels.mean() ))\n",
    "    print(\"Mean Absolute Error: \\t\", int(mean_absolute_error(test_labels, predicted_labels) ))\n",
    "    print(\"Mean Squared Error: \", mean_squared_error(test_labels, predicted_labels) )\n",
    "    print(\"Root Mean Squared Error: \", mean_squared_error(test_labels, predicted_labels, squared = False) )\n",
    "    print(\"MAPE :\" + str(mean_absolute_percentage_error( test_labels, predicted_labels )))\n",
    "    print(\"R2: \" + str(r2_score(test_labels,predicted_labels)) + \"\\n\")\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "import xgboost as xg"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "9619\n",
      "Mean True Value: \t 94755134\n",
      "Mean Absolute Error: \t 18211293\n",
      "Mean Squared Error:  1337993203338080.0\n",
      "Root Mean Squared Error:  36578589.41154073\n",
      "MAPE :1.1491019309627386\n",
      "R2: 0.8924750506244415\n",
      "\n",
      "1814\n",
      "Mean True Value: \t 94867525\n",
      "Mean Absolute Error: \t 18006477\n",
      "Mean Squared Error:  1287426527591657.2\n",
      "Root Mean Squared Error:  35880726.40836104\n",
      "MAPE :1.6971146192431594\n",
      "R2: 0.8962156663502199\n",
      "\n",
      "3773\n",
      "Mean True Value: \t 95372623\n",
      "Mean Absolute Error: \t 18154015\n",
      "Mean Squared Error:  1286976474545982.8\n",
      "Root Mean Squared Error:  35874454.34492325\n",
      "MAPE :1.2867656879329907\n",
      "R2: 0.8977113895045059\n",
      "\n"
     ]
    }
   ],
   "source": [
    "#EXTREME GRADIENT BOOST\n",
    "\n",
    "for i in range(3):\n",
    "    rseed = random.randint(1,10000)\n",
    "    print(rseed)\n",
    "\n",
    "    train_data, test_data, train_labels, test_labels = train_test_split(f,t, test_size=t_size, random_state=rseed)\n",
    "    xgb_r = xg.XGBRegressor(n_estimators = 1000, seed = 123)\n",
    "    xgb_r.fit(train_data, train_labels)\n",
    "    predicted_labels = xgb_r.predict(test_data)\n",
    "\n",
    "    print(\"Mean True Value: \\t\",int(test_labels.mean() ))\n",
    "    print(\"Mean Absolute Error: \\t\", int(mean_absolute_error(test_labels, predicted_labels) ))\n",
    "    print(\"Mean Squared Error: \", mean_squared_error(test_labels, predicted_labels) )\n",
    "    print(\"Root Mean Squared Error: \", mean_squared_error(test_labels, predicted_labels, squared = False) )\n",
    "    print(\"MAPE :\" + str(mean_absolute_percentage_error( test_labels, predicted_labels )))\n",
    "    print(\"R2: \" + str(r2_score(test_labels,predicted_labels)) + \"\\n\")\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1805\n",
      "Mean True Value: \t 94966868\n",
      "Mean Absolute Error: \t 33563042\n",
      "Mean Squared Error:  3393765972370551.5\n",
      "Root Mean Squared Error:  58256038.07649943\n",
      "MAPE :37.33825849381807\n",
      "R2: 0.7258908421466567\n",
      "\n",
      "3906\n",
      "Mean True Value: \t 95069446\n",
      "Mean Absolute Error: \t 33412157\n",
      "Mean Squared Error:  3300764536716695.0\n",
      "Root Mean Squared Error:  57452280.51798027\n",
      "MAPE :24.280783311141306\n",
      "R2: 0.7385441586149623\n",
      "\n",
      "7585\n",
      "Mean True Value: \t 94563971\n",
      "Mean Absolute Error: \t 32970799\n",
      "Mean Squared Error:  3188301303149065.5\n",
      "Root Mean Squared Error:  56465044.96721016\n",
      "MAPE :9.353637177832688\n",
      "R2: 0.7458677521565487\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.tree import DecisionTreeRegressor\n",
    "\n",
    "for i in range(3):\n",
    "    rseed = random.randint(1,10000)\n",
    "    print(rseed)\n",
    "\n",
    "    train_data, test_data, train_labels, test_labels = train_test_split(f,t, test_size= t_size, random_state=rseed)\n",
    "    \n",
    "    reg = DecisionTreeRegressor(max_depth = 4)   \n",
    "    \n",
    "\n",
    "    reg.fit(train_data, train_labels)\n",
    "    predicted_labels = reg.predict(test_data)\n",
    "\n",
    "\n",
    "    print(\"Mean True Value: \\t\",int(test_labels.mean() ))\n",
    "    print(\"Mean Absolute Error: \\t\", int(mean_absolute_error(test_labels, predicted_labels) ))\n",
    "    print(\"Mean Squared Error: \", mean_squared_error(test_labels, predicted_labels) )\n",
    "    print(\"Root Mean Squared Error: \", mean_squared_error(test_labels, predicted_labels, squared = False) )\n",
    "    print(\"MAPE :\" + str(mean_absolute_percentage_error( test_labels, predicted_labels )))\n",
    "    print(\"R2: \" + str(r2_score(test_labels,predicted_labels)) + \"\\n\")\n",
    "    \n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "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": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}