{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# TED-Seg Raw Data — Paragraph Statistics\n",
    "\n",
    "Inspect the paragraphs in `data/tedseg/raw/{train,val,test}.jsonl`.  \n",
    "Each item's `text` field is a list of paragraphs."
   ]
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-23T21:57:27.738194Z",
     "start_time": "2026-03-23T21:57:26.848369Z"
    }
   },
   "source": "import json\nimport re\nimport sys\nfrom pathlib import Path\nimport numpy as np\nimport matplotlib.pyplot as plt\n\nROOT = Path(__file__).resolve().parents[2] if \"__file__\" in dir() else Path.cwd().parents[1]\nsys.path.insert(0, str(ROOT))\n\nRAW_DIR = ROOT / \"data\" / \"tedseg\" / \"raw\"\nSPLITS = [\"train\", \"val\", \"test\"]",
   "outputs": [],
   "execution_count": 1
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Load all paragraphs per split"
   ]
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-23T21:57:30.131898Z",
     "start_time": "2026-03-23T21:57:27.758490Z"
    }
   },
   "source": [
    "split_data = {}  # split -> list of docs, each doc is list of paragraphs\n",
    "\n",
    "for split in SPLITS:\n",
    "    docs = []\n",
    "    with open(RAW_DIR / f\"{split}.jsonl\") as f:\n",
    "        for line in f:\n",
    "            line = line.strip()\n",
    "            if not line:\n",
    "                continue\n",
    "            obj = json.loads(line)\n",
    "            docs.append(obj[\"text\"])  # list of paragraph strings\n",
    "    split_data[split] = docs\n",
    "    print(f\"{split:>5}: {len(docs):,} documents\")"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "train: 4,154 documents\n",
      "  val: 519 documents\n",
      " test: 520 documents\n"
     ]
    }
   ],
   "execution_count": 2
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Per-split summary table"
   ]
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-23T21:57:35.514245Z",
     "start_time": "2026-03-23T21:57:33.316870Z"
    }
   },
   "source": [
    "print(f\"{'Split':>6} | {'Docs':>6} | {'Paragraphs':>11} | {'Paras/Doc':>12} | {'Words/Para':>12} | {'Chars/Para':>12} | {'Sents/Para (approx)':>20}\")\n",
    "print(\"-\" * 100)\n",
    "\n",
    "all_paragraphs = {}  # split -> flat list of paragraph strings\n",
    "\n",
    "for split in SPLITS:\n",
    "    docs = split_data[split]\n",
    "    paras = [p for doc in docs for p in doc]\n",
    "    all_paragraphs[split] = paras\n",
    "\n",
    "    n_docs = len(docs)\n",
    "    n_paras = len(paras)\n",
    "    paras_per_doc = [len(doc) for doc in docs]\n",
    "    word_counts = [len(p.split()) for p in paras]\n",
    "    char_counts = [len(p) for p in paras]\n",
    "    # rough sentence count by splitting on .!?\n",
    "    import re\n",
    "    sent_counts = [len([s for s in re.split(r'[.!?]+', p) if s.strip()]) for p in paras]\n",
    "\n",
    "    print(\n",
    "        f\"{split:>6} | {n_docs:>6,} | {n_paras:>11,} | \"\n",
    "        f\"{np.mean(paras_per_doc):>12.1f} | \"\n",
    "        f\"{np.mean(word_counts):>12.1f} | \"\n",
    "        f\"{np.mean(char_counts):>12.1f} | \"\n",
    "        f\"{np.mean(sent_counts):>20.1f}\"\n",
    "    )"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Split |   Docs |  Paragraphs |    Paras/Doc |   Words/Para |   Chars/Para |  Sents/Para (approx)\n",
      "----------------------------------------------------------------------------------------------------\n",
      " train |  4,154 |     104,425 |         25.1 |         68.9 |        386.8 |                  4.6\n",
      "   val |    519 |      13,426 |         25.9 |         68.6 |        384.1 |                  4.6\n",
      "  test |    520 |      13,224 |         25.4 |         69.5 |        388.9 |                  4.7\n"
     ]
    }
   ],
   "execution_count": 3
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Detailed paragraph statistics (all splits combined)"
   ]
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-23T21:57:37.607231Z",
     "start_time": "2026-03-23T21:57:35.524192Z"
    }
   },
   "source": [
    "all_paras = [p for split in SPLITS for p in all_paragraphs[split]]\n",
    "\n",
    "word_counts = np.array([len(p.split()) for p in all_paras])\n",
    "char_counts = np.array([len(p) for p in all_paras])\n",
    "sent_counts = np.array([len([s for s in re.split(r'[.!?]+', p) if s.strip()]) for p in all_paras])\n",
    "\n",
    "def print_stats(name, arr):\n",
    "    print(f\"\\n--- {name} ---\")\n",
    "    print(f\"  Total paragraphs : {len(arr):,}\")\n",
    "    print(f\"  Mean             : {np.mean(arr):.1f}\")\n",
    "    print(f\"  Std              : {np.std(arr):.1f}\")\n",
    "    print(f\"  Min              : {np.min(arr)}\")\n",
    "    print(f\"  25th percentile  : {np.percentile(arr, 25):.0f}\")\n",
    "    print(f\"  Median           : {np.median(arr):.0f}\")\n",
    "    print(f\"  75th percentile  : {np.percentile(arr, 75):.0f}\")\n",
    "    print(f\"  90th percentile  : {np.percentile(arr, 90):.0f}\")\n",
    "    print(f\"  95th percentile  : {np.percentile(arr, 95):.0f}\")\n",
    "    print(f\"  99th percentile  : {np.percentile(arr, 99):.0f}\")\n",
    "    print(f\"  Max              : {np.max(arr)}\")\n",
    "\n",
    "print_stats(\"Word count per paragraph\", word_counts)\n",
    "print_stats(\"Character count per paragraph\", char_counts)\n",
    "print_stats(\"Sentence count per paragraph (approx)\", sent_counts)"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "--- Word count per paragraph ---\n",
      "  Total paragraphs : 131,075\n",
      "  Mean             : 69.0\n",
      "  Std              : 67.6\n",
      "  Min              : 1\n",
      "  25th percentile  : 9\n",
      "  Median           : 58\n",
      "  75th percentile  : 106\n",
      "  90th percentile  : 153\n",
      "  95th percentile  : 187\n",
      "  99th percentile  : 266\n",
      "  Max              : 2744\n",
      "\n",
      "--- Character count per paragraph ---\n",
      "  Total paragraphs : 131,075\n",
      "  Mean             : 386.7\n",
      "  Std              : 377.8\n",
      "  Min              : 1\n",
      "  25th percentile  : 49\n",
      "  Median           : 328\n",
      "  75th percentile  : 594\n",
      "  90th percentile  : 859\n",
      "  95th percentile  : 1045\n",
      "  99th percentile  : 1481\n",
      "  Max              : 15074\n",
      "\n",
      "--- Sentence count per paragraph (approx) ---\n",
      "  Total paragraphs : 131,075\n",
      "  Mean             : 4.6\n",
      "  Std              : 4.2\n",
      "  Min              : 1\n",
      "  25th percentile  : 1\n",
      "  Median           : 4\n",
      "  75th percentile  : 6\n",
      "  90th percentile  : 10\n",
      "  95th percentile  : 12\n",
      "  99th percentile  : 18\n",
      "  Max              : 225\n"
     ]
    }
   ],
   "execution_count": 4
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. Paragraphs per document"
   ]
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-23T21:57:37.671722Z",
     "start_time": "2026-03-23T21:57:37.661765Z"
    }
   },
   "source": [
    "all_docs_para_counts = np.array([len(doc) for split in SPLITS for doc in split_data[split]])\n",
    "print_stats(\"Paragraphs per document\", all_docs_para_counts)"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "--- Paragraphs per document ---\n",
      "  Total paragraphs : 5,193\n",
      "  Mean             : 25.2\n",
      "  Std              : 21.1\n",
      "  Min              : 0\n",
      "  25th percentile  : 13\n",
      "  Median           : 20\n",
      "  75th percentile  : 30\n",
      "  90th percentile  : 45\n",
      "  95th percentile  : 60\n",
      "  99th percentile  : 107\n",
      "  Max              : 275\n"
     ]
    }
   ],
   "execution_count": 5
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. Histograms"
   ]
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-23T21:57:39.463197Z",
     "start_time": "2026-03-23T21:57:37.756272Z"
    }
   },
   "source": [
    "fig, axes = plt.subplots(2, 2, figsize=(14, 10))\n",
    "\n",
    "axes[0, 0].hist(word_counts, bins=80, edgecolor=\"black\", alpha=0.7)\n",
    "axes[0, 0].set_title(\"Word count per paragraph\")\n",
    "axes[0, 0].set_xlabel(\"Words\")\n",
    "axes[0, 0].set_ylabel(\"Count\")\n",
    "axes[0, 0].axvline(np.median(word_counts), color=\"red\", linestyle=\"--\", label=f\"Median={np.median(word_counts):.0f}\")\n",
    "axes[0, 0].legend()\n",
    "\n",
    "axes[0, 1].hist(char_counts, bins=80, edgecolor=\"black\", alpha=0.7)\n",
    "axes[0, 1].set_title(\"Character count per paragraph\")\n",
    "axes[0, 1].set_xlabel(\"Characters\")\n",
    "axes[0, 1].set_ylabel(\"Count\")\n",
    "axes[0, 1].axvline(np.median(char_counts), color=\"red\", linestyle=\"--\", label=f\"Median={np.median(char_counts):.0f}\")\n",
    "axes[0, 1].legend()\n",
    "\n",
    "axes[1, 0].hist(sent_counts, bins=range(0, max(sent_counts) + 2), edgecolor=\"black\", alpha=0.7)\n",
    "axes[1, 0].set_title(\"Sentence count per paragraph (approx)\")\n",
    "axes[1, 0].set_xlabel(\"Sentences\")\n",
    "axes[1, 0].set_ylabel(\"Count\")\n",
    "axes[1, 0].axvline(np.median(sent_counts), color=\"red\", linestyle=\"--\", label=f\"Median={np.median(sent_counts):.0f}\")\n",
    "axes[1, 0].legend()\n",
    "\n",
    "axes[1, 1].hist(all_docs_para_counts, bins=range(0, max(all_docs_para_counts) + 2), edgecolor=\"black\", alpha=0.7)\n",
    "axes[1, 1].set_title(\"Paragraphs per document\")\n",
    "axes[1, 1].set_xlabel(\"Paragraphs\")\n",
    "axes[1, 1].set_ylabel(\"Count\")\n",
    "axes[1, 1].axvline(np.median(all_docs_para_counts), color=\"red\", linestyle=\"--\", label=f\"Median={np.median(all_docs_para_counts):.0f}\")\n",
    "axes[1, 1].legend()\n",
    "\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ],
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 1400x1000 with 4 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data",
     "jetTransient": {
      "display_id": null
     }
    }
   ],
   "execution_count": 6
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 6. Per-split histograms (word count)"
   ]
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-23T21:57:47.653753Z",
     "start_time": "2026-03-23T21:57:44.072522Z"
    }
   },
   "source": "fig, axes = plt.subplots(1, 3, figsize=(18, 5), sharey=True)\n\nsplit_buckets = {}\n\nfor ax, split in zip(axes, SPLITS):\n    wc = [len(p.split()) for p in all_paragraphs[split]]\n    counts, edges, patches = ax.hist(wc, bins=60, edgecolor=\"black\", alpha=0.7)\n    ax.set_title(f\"{split} — words per paragraph (n={len(wc):,})\")\n    ax.set_xlabel(\"Words\")\n    ax.axvline(np.median(wc), color=\"red\", linestyle=\"--\", label=f\"Median={np.median(wc):.0f}\")\n    ax.legend()\n    split_buckets[split] = (counts, edges)\n\naxes[0].set_ylabel(\"Count\")\nplt.tight_layout()\nplt.show()\n\n# print bucket counts as text\nfor split in SPLITS:\n    counts, edges = split_buckets[split]\n    print(f\"\\n{'=' * 50}\")\n    print(f\"  {split.upper()} — word count bucket counts\")\n    print(f\"{'=' * 50}\")\n    print(f\"  {'Bucket range':>20}  |  {'Count':>7}\")\n    print(f\"  {'-'*20}--+--{'-'*7}\")\n    for i, count in enumerate(counts):\n        if count > 0:\n            print(f\"  {edges[i]:>8.0f} – {edges[i+1]:>8.0f}  |  {int(count):>7,}\")\n\n# per-split percentiles: word count\npcts = [5, 10, 25, 50, 75, 90, 95, 99]\nheader = f\"  {'Percentile':>10}\" + \"\".join(f\" | {s:>8}\" for s in SPLITS)\nsep = f\"  {'-'*10}\" + \"\".join(f\"-+-{'-'*8}\" for _ in SPLITS)\n\nprint(f\"\\n\\n{'=' * 50}\")\nprint(f\"  Per-split WORD COUNT percentiles\")\nprint(f\"{'=' * 50}\")\nprint(header)\nprint(sep)\nsplit_wc = {s: np.array([len(p.split()) for p in all_paragraphs[s]]) for s in SPLITS}\nfor p in pcts:\n    row = f\"  {p:>9}th\"\n    for s in SPLITS:\n        row += f\" | {np.percentile(split_wc[s], p):>8.0f}\"\n    print(row)\n\n# per-split percentiles: paragraph count per document\nprint(f\"\\n\\n{'=' * 50}\")\nprint(f\"  Per-split PARAGRAPHS PER DOCUMENT percentiles\")\nprint(f\"{'=' * 50}\")\nprint(header)\nprint(sep)\nsplit_pc = {s: np.array([len(doc) for doc in split_data[s]]) for s in SPLITS}\nfor p in pcts:\n    row = f\"  {p:>9}th\"\n    for s in SPLITS:\n        row += f\" | {np.percentile(split_pc[s], p):>8.0f}\"\n    print(row)",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 1800x500 with 3 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data",
     "jetTransient": {
      "display_id": null
     }
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "==================================================\n",
      "  TRAIN — word count bucket counts\n",
      "==================================================\n",
      "          Bucket range  |    Count\n",
      "  ----------------------+---------\n",
      "         1 –       47  |   45,172\n",
      "        47 –       92  |   26,644\n",
      "        92 –      138  |   18,488\n",
      "       138 –      184  |    8,538\n",
      "       184 –      230  |    3,456\n",
      "       230 –      275  |    1,230\n",
      "       275 –      321  |      464\n",
      "       321 –      367  |      202\n",
      "       367 –      412  |       93\n",
      "       412 –      458  |       69\n",
      "       458 –      504  |       19\n",
      "       504 –      550  |       11\n",
      "       550 –      595  |       11\n",
      "       595 –      641  |        7\n",
      "       641 –      687  |        5\n",
      "       687 –      732  |        2\n",
      "       732 –      778  |        1\n",
      "       824 –      870  |        3\n",
      "       870 –      915  |        1\n",
      "      1098 –     1144  |        1\n",
      "      1235 –     1281  |        1\n",
      "      1464 –     1510  |        2\n",
      "      1555 –     1601  |        1\n",
      "      1693 –     1738  |        1\n",
      "      2150 –     2195  |        1\n",
      "      2698 –     2744  |        2\n",
      "\n",
      "==================================================\n",
      "  VAL — word count bucket counts\n",
      "==================================================\n",
      "          Bucket range  |    Count\n",
      "  ----------------------+---------\n",
      "         1 –       44  |    5,768\n",
      "        44 –       87  |    3,197\n",
      "        87 –      130  |    2,346\n",
      "       130 –      173  |    1,210\n",
      "       173 –      216  |      528\n",
      "       216 –      258  |      202\n",
      "       258 –      301  |       95\n",
      "       301 –      344  |       29\n",
      "       344 –      387  |       21\n",
      "       387 –      430  |        5\n",
      "       430 –      473  |        5\n",
      "       473 –      516  |        5\n",
      "       516 –      559  |        2\n",
      "       559 –      602  |        3\n",
      "       644 –      687  |        2\n",
      "       687 –      730  |        1\n",
      "       730 –      773  |        2\n",
      "       859 –      902  |        1\n",
      "       945 –      988  |        1\n",
      "      1074 –     1116  |        1\n",
      "      1974 –     2017  |        1\n",
      "      2532 –     2575  |        1\n",
      "\n",
      "==================================================\n",
      "  TEST — word count bucket counts\n",
      "==================================================\n",
      "          Bucket range  |    Count\n",
      "  ----------------------+---------\n",
      "         1 –       44  |    5,482\n",
      "        44 –       87  |    3,166\n",
      "        87 –      130  |    2,426\n",
      "       130 –      173  |    1,234\n",
      "       173 –      216  |      550\n",
      "       216 –      259  |      240\n",
      "       259 –      302  |       74\n",
      "       302 –      345  |       26\n",
      "       345 –      388  |       10\n",
      "       388 –      431  |        6\n",
      "       431 –      474  |        4\n",
      "       474 –      517  |        1\n",
      "       517 –      560  |        2\n",
      "       560 –      603  |        1\n",
      "      1333 –     1376  |        1\n",
      "      2536 –     2579  |        1\n",
      "\n",
      "\n",
      "==================================================\n",
      "  Per-split WORD COUNT percentiles\n",
      "==================================================\n",
      "  Percentile |    train |      val |     test\n",
      "  -----------+----------+----------+---------\n",
      "          5th |        1 |        1 |        1\n",
      "         10th |        1 |        1 |        1\n",
      "         25th |        9 |       10 |       10\n",
      "         50th |       58 |       56 |       60\n",
      "         75th |      106 |      104 |      107\n",
      "         90th |      153 |      153 |      154\n",
      "         95th |      187 |      185 |      188\n",
      "         99th |      266 |      272 |      255\n",
      "\n",
      "\n",
      "==================================================\n",
      "  Per-split PARAGRAPHS PER DOCUMENT percentiles\n",
      "==================================================\n",
      "  Percentile |    train |      val |     test\n",
      "  -----------+----------+----------+---------\n",
      "          5th |        7 |        7 |        7\n",
      "         10th |        8 |        9 |        9\n",
      "         25th |       13 |       13 |       14\n",
      "         50th |       20 |       20 |       21\n",
      "         75th |       30 |       31 |       30\n",
      "         90th |       45 |       47 |       45\n",
      "         95th |       60 |       62 |       59\n",
      "         99th |      108 |      109 |       90\n"
     ]
    }
   ],
   "execution_count": 7
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 7. Shortest and longest paragraphs (examples)"
   ]
  },
  {
   "cell_type": "code",
   "source": "# compute 99th percentile word count per split, then take the mean\np99_per_split = []\nfor split in SPLITS:\n    wc = np.array([len(p.split()) for p in all_paragraphs[split]])\n    p99_per_split.append(np.percentile(wc, 99))\n\nthreshold = int(np.mean(p99_per_split))\nprint(f\"99th percentile word count per split: {dict(zip(SPLITS, [f'{v:.0f}' for v in p99_per_split]))}\")\nprint(f\"Mean of 99th percentiles (threshold): {threshold} words\\n\")\n\nprint(f\"{'Split':>6} | {'Total docs':>10} | {'Dropped':>8} | {'Kept':>8} | {'% dropped':>10}\")\nprint(\"-\" * 55)\n\nfor split in SPLITS:\n    docs = split_data[split]\n    total = len(docs)\n    dropped = sum(1 for doc in docs if any(len(p.split()) > threshold for p in doc))\n    kept = total - dropped\n    print(f\"{split:>6} | {total:>10,} | {dropped:>8,} | {kept:>8,} | {dropped/total*100:>9.1f}%\")\n\nall_total = sum(len(split_data[s]) for s in SPLITS)\nall_dropped = sum(\n    1 for s in SPLITS for doc in split_data[s]\n    if any(len(p.split()) > threshold for p in doc)\n)\nprint(f\"\\n{'ALL':>6} | {all_total:>10,} | {all_dropped:>8,} | {all_total - all_dropped:>8,} | {all_dropped/all_total*100:>9.1f}%\")",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-23T21:57:49.922599Z",
     "start_time": "2026-03-23T21:57:47.753473Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "99th percentile word count per split: {'train': '266', 'val': '272', 'test': '255'}\n",
      "Mean of 99th percentiles (threshold): 264 words\n",
      "\n",
      " Split | Total docs |  Dropped |     Kept |  % dropped\n",
      "-------------------------------------------------------\n",
      " train |      4,154 |      599 |    3,555 |      14.4%\n",
      "   val |        519 |       86 |      433 |      16.6%\n",
      "  test |        520 |       72 |      448 |      13.8%\n",
      "\n",
      "   ALL |      5,193 |      757 |    4,436 |      14.6%\n"
     ]
    }
   ],
   "execution_count": 8
  },
  {
   "cell_type": "markdown",
   "source": "## 6c. Impact of dropping docs with paragraphs longer than mean of 99th percentile word counts",
   "metadata": {}
  },
  {
   "cell_type": "code",
   "source": "# per-split word count stats\nfor split in SPLITS:\n    wc = np.array([len(p.split()) for p in all_paragraphs[split]])\n    print(f\"\\n{'=' * 50}\")\n    print(f\"  {split.upper()} — word count per paragraph\")\n    print(f\"{'=' * 50}\")\n    print_stats(f\"{split} word count\", wc)\n\n# per-split paragraph count stats\nfor split in SPLITS:\n    pc = np.array([len(doc) for doc in split_data[split]])\n    print(f\"\\n{'=' * 50}\")\n    print(f\"  {split.upper()} — paragraphs per document\")\n    print(f\"{'=' * 50}\")\n    print_stats(f\"{split} para count\", pc)",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-23T21:57:50.724218Z",
     "start_time": "2026-03-23T21:57:49.991714Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "==================================================\n",
      "  TRAIN — word count per paragraph\n",
      "==================================================\n",
      "\n",
      "--- train word count ---\n",
      "  Total paragraphs : 104,425\n",
      "  Mean             : 68.9\n",
      "  Std              : 67.0\n",
      "  Min              : 1\n",
      "  25th percentile  : 9\n",
      "  Median           : 58\n",
      "  75th percentile  : 106\n",
      "  90th percentile  : 153\n",
      "  95th percentile  : 187\n",
      "  99th percentile  : 266\n",
      "  Max              : 2744\n",
      "\n",
      "==================================================\n",
      "  VAL — word count per paragraph\n",
      "==================================================\n",
      "\n",
      "--- val word count ---\n",
      "  Total paragraphs : 13,426\n",
      "  Mean             : 68.6\n",
      "  Std              : 72.3\n",
      "  Min              : 1\n",
      "  25th percentile  : 10\n",
      "  Median           : 56\n",
      "  75th percentile  : 104\n",
      "  90th percentile  : 153\n",
      "  95th percentile  : 185\n",
      "  99th percentile  : 272\n",
      "  Max              : 2575\n",
      "\n",
      "==================================================\n",
      "  TEST — word count per paragraph\n",
      "==================================================\n",
      "\n",
      "--- test word count ---\n",
      "  Total paragraphs : 13,224\n",
      "  Mean             : 69.5\n",
      "  Std              : 68.1\n",
      "  Min              : 1\n",
      "  25th percentile  : 10\n",
      "  Median           : 60\n",
      "  75th percentile  : 107\n",
      "  90th percentile  : 154\n",
      "  95th percentile  : 188\n",
      "  99th percentile  : 255\n",
      "  Max              : 2579\n",
      "\n",
      "==================================================\n",
      "  TRAIN — paragraphs per document\n",
      "==================================================\n",
      "\n",
      "--- train para count ---\n",
      "  Total paragraphs : 4,154\n",
      "  Mean             : 25.1\n",
      "  Std              : 20.8\n",
      "  Min              : 1\n",
      "  25th percentile  : 13\n",
      "  Median           : 20\n",
      "  75th percentile  : 30\n",
      "  90th percentile  : 45\n",
      "  95th percentile  : 60\n",
      "  99th percentile  : 108\n",
      "  Max              : 275\n",
      "\n",
      "==================================================\n",
      "  VAL — paragraphs per document\n",
      "==================================================\n",
      "\n",
      "--- val para count ---\n",
      "  Total paragraphs : 519\n",
      "  Mean             : 25.9\n",
      "  Std              : 23.6\n",
      "  Min              : 2\n",
      "  25th percentile  : 13\n",
      "  Median           : 20\n",
      "  75th percentile  : 31\n",
      "  90th percentile  : 47\n",
      "  95th percentile  : 62\n",
      "  99th percentile  : 109\n",
      "  Max              : 222\n",
      "\n",
      "==================================================\n",
      "  TEST — paragraphs per document\n",
      "==================================================\n",
      "\n",
      "--- test para count ---\n",
      "  Total paragraphs : 520\n",
      "  Mean             : 25.4\n",
      "  Std              : 20.6\n",
      "  Min              : 0\n",
      "  25th percentile  : 14\n",
      "  Median           : 21\n",
      "  75th percentile  : 30\n",
      "  90th percentile  : 45\n",
      "  95th percentile  : 59\n",
      "  99th percentile  : 90\n",
      "  Max              : 255\n"
     ]
    }
   ],
   "execution_count": 9
  },
  {
   "cell_type": "markdown",
   "source": "## 6b. Per-split detailed statistics (word count & paragraph count)",
   "metadata": {}
  },
  {
   "cell_type": "code",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2026-03-23T21:57:51.454700Z",
     "start_time": "2026-03-23T21:57:50.741527Z"
    }
   },
   "source": [
    "sorted_by_words = sorted(all_paras, key=lambda p: len(p.split()))\n",
    "\n",
    "print(\"=== 5 SHORTEST paragraphs (by word count) ===\")\n",
    "for p in sorted_by_words[:5]:\n",
    "    print(f\"  [{len(p.split()):>3} words] {p[:150]}{'...' if len(p) > 150 else ''}\")\n",
    "\n",
    "print(f\"\\n=== 5 LONGEST paragraphs (by word count) ===\")\n",
    "for p in sorted_by_words[-5:]:\n",
    "    print(f\"  [{len(p.split()):>3} words] {p[:150]}{'...' if len(p) > 150 else ''}\")"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "=== 5 SHORTEST paragraphs (by word count) ===\n",
      "  [  1 words] (Laughter)\n",
      "  [  1 words] (Music)\n",
      "  [  1 words] (Music)\n",
      "  [  1 words] upbeat\n",
      "  [  1 words] (Music)\n",
      "\n",
      "=== 5 LONGEST paragraphs (by word count) ===\n",
      "  [2160 words] And looking at that photograph, I began to feel nauseous. I thought I might throw up into my screen, and maybe it was the vodka. But I think it was ac...\n",
      "  [2575 words] don't know how to read nor write at all, yet. Over there are those who know, or say they know,  how to read and write. So, what do we have? We have th...\n",
      "  [2579 words] Who can we trust to tell us the truth? Once in another tumultuous era, as the Vietnam War, Watergate, the Civil Rights and Women's Liberation Movement...\n",
      "  [2728 words] But some of us might've smelt something a little bit less pleasant, perhaps - perhaps somebody's bad breath or body odour. Maybe you even smelled your...\n",
      "  [2744 words] With Cliteracy I felt that - Yes, language has been this way of restricting and confining the female body, but if language can do these thing, it can ...\n"
     ]
    }
   ],
   "execution_count": 10
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "name": "python",
   "version": "3.10.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}