xplainproj/xplain_dataset_prep.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "id": "ee95a38117f298b2",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2025-02-08T20:56:44.869721Z",
     "start_time": "2025-02-08T20:56:44.866338Z"
    }
   },
   "source": [
    "from IPython.display import display, HTML\n",
    "\n",
    "# Set the width to a specific percentage of the screen width, e.g., 100%\n",
    "display(HTML('<style>.container { width:100% !important; }</style>'))"
   ],
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ],
      "text/html": [
       "<style>.container { width:100% !important; }</style>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "execution_count": 1
  },
  {
   "cell_type": "code",
   "id": "initial_id",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2025-02-08T20:56:44.874929Z",
     "start_time": "2025-02-08T20:56:44.873142Z"
    }
   },
   "source": [
    "from os import path\n",
    "\n",
    "PROJECT_ROOT = path.abspath(path.join(globals()['_dh'][0], '..'))\n",
    "DATALAKE_PATH = path.abspath(path.join(PROJECT_ROOT, '..', '..', 'datalake', 'xplainproj'))\n",
    "FQFN_PROCESSED_DF = path.join(DATALAKE_PATH, 'processed', f'processed_20250207.jsonl')"
   ],
   "outputs": [],
   "execution_count": 2
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2025-02-08T20:56:56.016482Z",
     "start_time": "2025-02-08T20:56:44.924699Z"
    }
   },
   "cell_type": "code",
   "source": [
    "from xplainproj.trainer import xplain_folder_loader\n",
    "\n",
    "wtl = xplain_folder_loader.XplainFolderLoader()\n",
    "wtl.read(DATALAKE_PATH, labels=['0', '1'])"
   ],
   "id": "b7d79f6ed3a4a866",
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2025-02-08 12:56:46,527 - tensorcraft - INFO - XLA Device Not Supported: No module named 'torch_xla'\n",
      "2025-02-08 12:56:46,536 - tensorcraft - INFO - Pytorch version=2.6.0 preferred device=mps build with MPS support=True\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Creating tasks:   0%|          | 0/2 [00:00<?, ?it/s]"
      ],
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "395d0ab09bca417f9b15b9d6b13f4b4f"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "Test files processing:   0%|          | 0/704 [00:00<?, ?it/s]"
      ],
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "7b4d55c75a8448218c459cda0d8a66d8"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2025-02-08 12:56:48,802 - tensorcraft - INFO - XLA Device Not Supported: No module named 'torch_xla'\n",
      "2025-02-08 12:56:48,817 - tensorcraft - INFO - Pytorch version=2.6.0 preferred device=mps build with MPS support=True\n",
      "2025-02-08 12:56:48,870 - tensorcraft - INFO - XLA Device Not Supported: No module named 'torch_xla'\n",
      "2025-02-08 12:56:48,871 - tensorcraft - INFO - XLA Device Not Supported: No module named 'torch_xla'\n",
      "2025-02-08 12:56:48,883 - tensorcraft - INFO - Pytorch version=2.6.0 preferred device=mps build with MPS support=True\n",
      "2025-02-08 12:56:48,884 - tensorcraft - INFO - Pytorch version=2.6.0 preferred device=mps build with MPS support=True\n",
      "2025-02-08 12:56:48,894 - tensorcraft - INFO - XLA Device Not Supported: No module named 'torch_xla'\n",
      "2025-02-08 12:56:48,909 - tensorcraft - INFO - Pytorch version=2.6.0 preferred device=mps build with MPS support=True\n",
      "2025-02-08 12:56:48,919 - tensorcraft - INFO - XLA Device Not Supported: No module named 'torch_xla'\n",
      "2025-02-08 12:56:48,933 - tensorcraft - INFO - Pytorch version=2.6.0 preferred device=mps build with MPS support=True\n",
      "2025-02-08 12:56:48,954 - tensorcraft - INFO - XLA Device Not Supported: No module named 'torch_xla'\n",
      "2025-02-08 12:56:48,957 - tensorcraft - INFO - XLA Device Not Supported: No module named 'torch_xla'\n",
      "2025-02-08 12:56:48,959 - tensorcraft - INFO - XLA Device Not Supported: No module named 'torch_xla'\n",
      "2025-02-08 12:56:48,966 - tensorcraft - INFO - Pytorch version=2.6.0 preferred device=mps build with MPS support=True\n",
      "2025-02-08 12:56:48,967 - tensorcraft - INFO - XLA Device Not Supported: No module named 'torch_xla'\n",
      "2025-02-08 12:56:48,969 - tensorcraft - INFO - Pytorch version=2.6.0 preferred device=mps build with MPS support=True\n",
      "2025-02-08 12:56:48,970 - tensorcraft - INFO - Pytorch version=2.6.0 preferred device=mps build with MPS support=True\n",
      "2025-02-08 12:56:48,979 - tensorcraft - INFO - XLA Device Not Supported: No module named 'torch_xla'\n",
      "2025-02-08 12:56:48,979 - tensorcraft - INFO - XLA Device Not Supported: No module named 'torch_xla'\n",
      "2025-02-08 12:56:48,980 - tensorcraft - INFO - Pytorch version=2.6.0 preferred device=mps build with MPS support=True\n",
      "2025-02-08 12:56:48,992 - tensorcraft - INFO - Pytorch version=2.6.0 preferred device=mps build with MPS support=True\n",
      "2025-02-08 12:56:48,993 - tensorcraft - INFO - Pytorch version=2.6.0 preferred device=mps build with MPS support=True\n",
      "2025-02-08 12:56:49,004 - tensorcraft - INFO - XLA Device Not Supported: No module named 'torch_xla'\n",
      "2025-02-08 12:56:49,028 - tensorcraft - INFO - XLA Device Not Supported: No module named 'torch_xla'\n",
      "2025-02-08 12:56:49,036 - tensorcraft - INFO - Pytorch version=2.6.0 preferred device=mps build with MPS support=True\n",
      "2025-02-08 12:56:49,048 - tensorcraft - INFO - XLA Device Not Supported: No module named 'torch_xla'\n",
      "2025-02-08 12:56:49,048 - tensorcraft - INFO - Pytorch version=2.6.0 preferred device=mps build with MPS support=True\n",
      "2025-02-08 12:56:49,066 - tensorcraft - INFO - Pytorch version=2.6.0 preferred device=mps build with MPS support=True\n"
     ]
    }
   ],
   "execution_count": 3
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2025-02-08T20:56:56.029742Z",
     "start_time": "2025-02-08T20:56:56.020592Z"
    }
   },
   "cell_type": "code",
   "source": [
    "df = wtl.df\n",
    "df_filtered = df.loc[:, df.columns != 'fqfn']\n",
    "df_filtered"
   ],
   "id": "ce60ca813b9b4562",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "    file_name                                          text_body  label  \\\n",
       "0    1289.ps1  \\nfunction Get-CMsmqMessageQueue\\n{\\n    \\n   ...      0   \n",
       "1    1061.ps1  \\n\\n\\n\\n\\n\\n\\n\\n\\n\\n\\n\\nSet-StrictMode -Versio...      0   \n",
       "2    1262.ps1  \\nfunction Test-CDscTargetResource\\n{\\n    \\n ...      0   \n",
       "3    1049.ps1  \\n\\n\\n\\n\\n\\n\\n\\n\\n\\n\\n\\n& (Join-Path -Path $PS...      0   \n",
       "4    1263.ps1  \\nfunction Test-CFirewallStatefulFtp\\n{\\n    \\...      0   \n",
       "..        ...                                                ...    ...   \n",
       "699   894.ps1  \\n$BHgK = '[DllImport(\"kernel32.dll\")]public s...      1   \n",
       "700  2809.ps1  \\n((new-object net.webclient).DownloadFile('ht...      1   \n",
       "701   102.ps1  \\n$c = '[DllImport(\"kernel32.dll\")]public stat...      1   \n",
       "702  2821.ps1  \\n$1 = '$c = ''[DllImport(\"kernel32.dll\")]publ...      1   \n",
       "703   116.ps1  \\n$c = '[DllImport(\"kernel32.dll\")]public stat...      1   \n",
       "\n",
       "     longest_code_line_length  median_code_line_length  lines_of_code  \\\n",
       "0                         1.0                      1.0       9.312883   \n",
       "1                         1.0                      1.0      11.388017   \n",
       "2                         1.0                      1.0      10.856426   \n",
       "3                         1.0                      1.0      10.653741   \n",
       "4                         1.0                      1.0       8.636625   \n",
       "..                        ...                      ...            ...   \n",
       "699                       1.0                      1.0      11.314583   \n",
       "700                       1.0                      1.0       7.179909   \n",
       "701                       1.0                      1.0      11.126059   \n",
       "702                       1.0                      1.0      11.279030   \n",
       "703                       1.0                      1.0      11.258566   \n",
       "\n",
       "     code_size_in_bytes  ratio_of_comments_to_code  is_64base_content_present  \\\n",
       "0              9.312883                      -0.11                      False   \n",
       "1             11.388017                      -0.09                      False   \n",
       "2             10.856426                      -0.10                      False   \n",
       "3             10.653741                      -0.10                      False   \n",
       "4              8.636625                      -0.12                      False   \n",
       "..                  ...                        ...                        ...   \n",
       "699           11.314583                      -0.09                      False   \n",
       "700            7.179909                      -0.14                      False   \n",
       "701           11.126059                      -0.09                      False   \n",
       "702           11.279030                      -0.09                      False   \n",
       "703           11.258566                      -0.09                      False   \n",
       "\n",
       "                                   file_name_embedding  \n",
       "0    [-0.68170124, -0.5240398, 0.14555997, 0.277613...  \n",
       "1    [-0.68159485, -0.48227602, 0.20190568, 0.29202...  \n",
       "2    [-0.68062484, -0.5328485, 0.15031643, 0.272234...  \n",
       "3    [-0.6629062, -0.4953126, 0.22535203, 0.2530482...  \n",
       "4    [-0.6847729, -0.52624404, 0.13542272, 0.289565...  \n",
       "..                                                 ...  \n",
       "699  [-0.7092016, -0.48694262, 0.3059109, 0.4413421...  \n",
       "700  [-0.7066325, -0.49039826, 0.24720317, 0.229103...  \n",
       "701  [-0.68433297, -0.49844956, 0.34662682, 0.34134...  \n",
       "702  [-0.7025869, -0.49556226, 0.25579977, 0.213931...  \n",
       "703  [-0.676445, -0.48426792, 0.3580541, 0.3581275,...  \n",
       "\n",
       "[704 rows x 10 columns]"
      ],
      "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>file_name</th>\n",
       "      <th>text_body</th>\n",
       "      <th>label</th>\n",
       "      <th>longest_code_line_length</th>\n",
       "      <th>median_code_line_length</th>\n",
       "      <th>lines_of_code</th>\n",
       "      <th>code_size_in_bytes</th>\n",
       "      <th>ratio_of_comments_to_code</th>\n",
       "      <th>is_64base_content_present</th>\n",
       "      <th>file_name_embedding</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1289.ps1</td>\n",
       "      <td>\\nfunction Get-CMsmqMessageQueue\\n{\\n    \\n   ...</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>9.312883</td>\n",
       "      <td>9.312883</td>\n",
       "      <td>-0.11</td>\n",
       "      <td>False</td>\n",
       "      <td>[-0.68170124, -0.5240398, 0.14555997, 0.277613...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1061.ps1</td>\n",
       "      <td>\\n\\n\\n\\n\\n\\n\\n\\n\\n\\n\\n\\nSet-StrictMode -Versio...</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>11.388017</td>\n",
       "      <td>11.388017</td>\n",
       "      <td>-0.09</td>\n",
       "      <td>False</td>\n",
       "      <td>[-0.68159485, -0.48227602, 0.20190568, 0.29202...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1262.ps1</td>\n",
       "      <td>\\nfunction Test-CDscTargetResource\\n{\\n    \\n ...</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>10.856426</td>\n",
       "      <td>10.856426</td>\n",
       "      <td>-0.10</td>\n",
       "      <td>False</td>\n",
       "      <td>[-0.68062484, -0.5328485, 0.15031643, 0.272234...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1049.ps1</td>\n",
       "      <td>\\n\\n\\n\\n\\n\\n\\n\\n\\n\\n\\n\\n&amp; (Join-Path -Path $PS...</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>10.653741</td>\n",
       "      <td>10.653741</td>\n",
       "      <td>-0.10</td>\n",
       "      <td>False</td>\n",
       "      <td>[-0.6629062, -0.4953126, 0.22535203, 0.2530482...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1263.ps1</td>\n",
       "      <td>\\nfunction Test-CFirewallStatefulFtp\\n{\\n    \\...</td>\n",
       "      <td>0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8.636625</td>\n",
       "      <td>8.636625</td>\n",
       "      <td>-0.12</td>\n",
       "      <td>False</td>\n",
       "      <td>[-0.6847729, -0.52624404, 0.13542272, 0.289565...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>699</th>\n",
       "      <td>894.ps1</td>\n",
       "      <td>\\n$BHgK = '[DllImport(\"kernel32.dll\")]public s...</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>11.314583</td>\n",
       "      <td>11.314583</td>\n",
       "      <td>-0.09</td>\n",
       "      <td>False</td>\n",
       "      <td>[-0.7092016, -0.48694262, 0.3059109, 0.4413421...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>700</th>\n",
       "      <td>2809.ps1</td>\n",
       "      <td>\\n((new-object net.webclient).DownloadFile('ht...</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>7.179909</td>\n",
       "      <td>7.179909</td>\n",
       "      <td>-0.14</td>\n",
       "      <td>False</td>\n",
       "      <td>[-0.7066325, -0.49039826, 0.24720317, 0.229103...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>701</th>\n",
       "      <td>102.ps1</td>\n",
       "      <td>\\n$c = '[DllImport(\"kernel32.dll\")]public stat...</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>11.126059</td>\n",
       "      <td>11.126059</td>\n",
       "      <td>-0.09</td>\n",
       "      <td>False</td>\n",
       "      <td>[-0.68433297, -0.49844956, 0.34662682, 0.34134...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>702</th>\n",
       "      <td>2821.ps1</td>\n",
       "      <td>\\n$1 = '$c = ''[DllImport(\"kernel32.dll\")]publ...</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>11.279030</td>\n",
       "      <td>11.279030</td>\n",
       "      <td>-0.09</td>\n",
       "      <td>False</td>\n",
       "      <td>[-0.7025869, -0.49556226, 0.25579977, 0.213931...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>703</th>\n",
       "      <td>116.ps1</td>\n",
       "      <td>\\n$c = '[DllImport(\"kernel32.dll\")]public stat...</td>\n",
       "      <td>1</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>11.258566</td>\n",
       "      <td>11.258566</td>\n",
       "      <td>-0.09</td>\n",
       "      <td>False</td>\n",
       "      <td>[-0.676445, -0.48426792, 0.3580541, 0.3581275,...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>704 rows × 10 columns</p>\n",
       "</div>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 4
  },
  {
   "cell_type": "code",
   "id": "7414ecfae34ebc92",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2025-02-08T20:56:56.208006Z",
     "start_time": "2025-02-08T20:56:56.046477Z"
    }
   },
   "source": [
    "df['label'].hist()"
   ],
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "execution_count": 5
  },
  {
   "cell_type": "code",
   "id": "319905f9545ff990",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2025-02-08T20:56:56.228221Z",
     "start_time": "2025-02-08T20:56:56.223294Z"
    }
   },
   "source": [
    "for label in df['label'].unique():\n",
    "    sampled_df = df[df['label'] == label]['file_name_embedding'].sample(n=10)\n",
    "    print(f'Sampled DataFrame for label {label}:')\n",
    "    display(sampled_df)"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sampled DataFrame for label 0:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "112    [-0.6448472, -0.5418714, 0.2353214, 0.26844865...\n",
       "166    [-0.6702795, -0.50405985, 0.17132682, 0.264245...\n",
       "141    [-0.66448843, -0.5018722, 0.18129341, 0.310334...\n",
       "94     [-0.67616785, -0.45102385, 0.3707155, 0.277383...\n",
       "285    [-0.6196193, -0.56098115, 0.2365901, 0.15419, ...\n",
       "190    [-0.63850677, -0.53774375, 0.24473329, 0.23865...\n",
       "162    [-0.61189085, -0.50936425, 0.22962943, 0.12033...\n",
       "243    [-0.65730363, -0.48580417, 0.2074899, 0.278497...\n",
       "95     [-0.54708135, -0.51365334, 0.2565986, 0.209010...\n",
       "122    [-0.6588425, -0.4890083, 0.20582764, 0.2679962...\n",
       "Name: file_name_embedding, dtype: object"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Sampled DataFrame for label 1:\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "384    [-0.70630074, -0.5597498, 0.16273329, 0.222851...\n",
       "402    [-0.7212416, -0.4711705, 0.32801896, 0.3409890...\n",
       "608    [-0.6794944, -0.5104539, 0.21766038, 0.2759730...\n",
       "568    [-0.631977, -0.51373994, 0.2293594, 0.26230276...\n",
       "366    [-0.68170124, -0.5240398, 0.14555997, 0.277613...\n",
       "536    [-0.66723865, -0.5226401, 0.16759437, 0.271389...\n",
       "614    [-0.7409419, -0.4822608, 0.3088669, 0.4314015,...\n",
       "606    [-0.6965097, -0.50748247, 0.19465749, 0.220671...\n",
       "458    [-0.7203547, -0.4971552, 0.17465828, 0.2785666...\n",
       "375    [-0.6629062, -0.4953126, 0.22535203, 0.2530482...\n",
       "Name: file_name_embedding, dtype: object"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "execution_count": 6
  },
  {
   "cell_type": "code",
   "id": "2c214ba4cf09217c",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2025-02-08T20:56:56.261783Z",
     "start_time": "2025-02-08T20:56:56.245176Z"
    }
   },
   "source": [
    "df.to_json(FQFN_PROCESSED_DF, orient='records', lines=True)"
   ],
   "outputs": [],
   "execution_count": 7
  },
  {
   "cell_type": "code",
   "id": "cf0a199ddc53bd4c",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2025-02-08T20:56:56.460820Z",
     "start_time": "2025-02-08T20:56:56.459553Z"
    }
   },
   "source": [],
   "outputs": [],
   "execution_count": null
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}