{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "import json\n",
    "import tempfile\n",
    "import os\n",
    "import boto3\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from tqdm import tqdm\n",
    "from itertools import combinations\n",
    "\n",
    "tqdm.pandas()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# DATA_PATH = \"/home/sara/glockenspiel/suno_utils/task_eval/modal_runs/genre_mappings_chirp-v5-sem-6bdiversity_test_2025_03_03-16_53_11.json\"\n",
    "# DATA_PATH = \"/home/sara/glockenspiel/suno_utils/task_eval/modal_runs/genre_mappings_chirp-v4-h-s-32diversity_test_2025_03_03-15_33_39.json\"\n",
    "# DATA_PATH = \"/home/sara/glockenspiel/suno_utils/task_eval/modal_runs/genre_mappings_chirp-v5-sem-6b_no_tag_cfg_diversity_2025_03_03-20_07_35.json\"\n",
    "# DATA_PATH = \"/home/sara/glockenspiel/suno_utils/task_eval/modal_runs/genre_mappings_chirp-v5-sem-6b-dpo_2025_03_03-21_28_14.json\"\n",
    "# DATA_PATH = \"/home/sara/glockenspiel/suno_utils/task_eval/modal_runs/genre_mappings_chirp-v5-sem-6b_base_genre_2025_03_03-23_38_03.json\"\n",
    "# DATA_PATH = \"/home/sara/glockenspiel/suno_utils/task_eval/modal_runs/genre_mappings_chirp-v4-h-s-32_base_genre_2025_03_04-02_29_27.json\"\n",
    "DATA_PATH = \"/home/sara/glockenspiel/suno_utils/task_eval/modal_runs/genre_mappings_chirp-auk-eval_blue_t1r7_2025_06_06-20_20_12.json\"\n",
    "# DATA_PATH = \"/home/sara/glockenspiel/suno_utils/task_eval/modal_runs/genre_mappings_chirp-auk-eval_426_2025_04_29-20_25_49.json\"\n",
    "TIMESTAMP = \"2025_06_06-20_20_12\"\n",
    "MODEL = \"BlueJay t1-r7\"\n",
    "\n",
    "DITTO_S3_PATH = \"s3://suno-data/minz/models/ditto_v2_epoch_57.pt\"\n",
    "TASK = \"self_sim\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_gen_json(filepath):\n",
    "    with open(filepath, \"r\", encoding=\"utf-8\") as file:\n",
    "        data = json.load(file)\n",
    "    by_item = []\n",
    "    for genre, outputs in data.items():\n",
    "        for val in outputs:\n",
    "            val[\"genre\"] = genre\n",
    "            by_item.append(val)\n",
    "            val[\"instrumental\"] = len(val[\"lyrics\"]) < 15\n",
    "            if val[\"instrumental\"]:\n",
    "                val[\"lyrics\"] = [\"Instrumental\"]\n",
    "    df = pd.DataFrame(by_item)\n",
    "\n",
    "    return df\n",
    "\n",
    "\n",
    "def cosine_similarity(a, b):\n",
    "    dot_product = np.dot(a, b)\n",
    "    magnitude_a = np.sqrt(np.dot(a, a))\n",
    "    magnitude_b = np.sqrt(np.dot(b, b))\n",
    "    return dot_product / (magnitude_a * magnitude_b)\n",
    "\n",
    "\n",
    "def get_task_score(row, task=TASK):\n",
    "    s3 = boto3.client(\"s3\")\n",
    "    bucket = \"suno-data-uploads\"\n",
    "    folder = f\"tasks/feature_eval/cover_persona/{TIMESTAMP}/\"\n",
    "    s3_id = row[\"s3_id\"]\n",
    "\n",
    "    with tempfile.NamedTemporaryFile(suffix=\".npz\") as temp_file:\n",
    "        try:\n",
    "            s3.download_file(bucket, os.path.join(folder, f\"{s3_id}_{task}_ditto.npz\"), temp_file.name)\n",
    "            data = np.load(temp_file.name)\n",
    "            source_task = data[\"task\"]\n",
    "            source_ditto = data[\"embedding\"]\n",
    "            assert source_task == task\n",
    "        except Exception as e:\n",
    "            print(f\"Skipping? {e}\")\n",
    "            return None\n",
    "\n",
    "    return source_ditto"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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>s3_id</th>\n",
       "      <th>tags</th>\n",
       "      <th>lyrics</th>\n",
       "      <th>genre</th>\n",
       "      <th>instrumental</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1252d952-dd55-4fcc-88d6-1efea11d13db</td>\n",
       "      <td>808, smooth, r&amp;b, bass, soul, blues, male voca...</td>\n",
       "      <td>(Verse)\\nDancing so close, I can feel your bod...</td>\n",
       "      <td>r&amp;b</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4e7aaa68-8bd1-48e1-8f5f-585214a202bf</td>\n",
       "      <td>808, smooth, r&amp;b, bass, soul, blues, male voca...</td>\n",
       "      <td>(Verse)\\nDancing so close, I can feel your bod...</td>\n",
       "      <td>r&amp;b</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0219984d-4e98-4279-b549-ce8cbad00cb5</td>\n",
       "      <td>808, smooth, r&amp;b, bass, soul, blues, male voca...</td>\n",
       "      <td>(Verse)\\nDancing so close, I can feel your bod...</td>\n",
       "      <td>r&amp;b</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>70bc6079-7e32-48f2-ab8b-fba02669fe9a</td>\n",
       "      <td>808, smooth, r&amp;b, bass, soul, blues, male voca...</td>\n",
       "      <td>(Verse)\\nDancing so close, I can feel your bod...</td>\n",
       "      <td>r&amp;b</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>74326bb6-56bb-4b47-98cd-03f74091b85d</td>\n",
       "      <td>808, smooth, r&amp;b, bass, soul, blues, male voca...</td>\n",
       "      <td>(Verse)\\nDancing so close, I can feel your bod...</td>\n",
       "      <td>r&amp;b</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  s3_id  \\\n",
       "0  1252d952-dd55-4fcc-88d6-1efea11d13db   \n",
       "1  4e7aaa68-8bd1-48e1-8f5f-585214a202bf   \n",
       "2  0219984d-4e98-4279-b549-ce8cbad00cb5   \n",
       "3  70bc6079-7e32-48f2-ab8b-fba02669fe9a   \n",
       "4  74326bb6-56bb-4b47-98cd-03f74091b85d   \n",
       "\n",
       "                                                tags  \\\n",
       "0  808, smooth, r&b, bass, soul, blues, male voca...   \n",
       "1  808, smooth, r&b, bass, soul, blues, male voca...   \n",
       "2  808, smooth, r&b, bass, soul, blues, male voca...   \n",
       "3  808, smooth, r&b, bass, soul, blues, male voca...   \n",
       "4  808, smooth, r&b, bass, soul, blues, male voca...   \n",
       "\n",
       "                                              lyrics genre  instrumental  \n",
       "0  (Verse)\\nDancing so close, I can feel your bod...   r&b         False  \n",
       "1  (Verse)\\nDancing so close, I can feel your bod...   r&b         False  \n",
       "2  (Verse)\\nDancing so close, I can feel your bod...   r&b         False  \n",
       "3  (Verse)\\nDancing so close, I can feel your bod...   r&b         False  \n",
       "4  (Verse)\\nDancing so close, I can feel your bod...   r&b         False  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gen_df = load_gen_json(DATA_PATH)\n",
    "gen_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1500/1500 [04:07<00:00,  6.07it/s]\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>s3_id</th>\n",
       "      <th>tags</th>\n",
       "      <th>lyrics</th>\n",
       "      <th>genre</th>\n",
       "      <th>instrumental</th>\n",
       "      <th>ditto_embed</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1252d952-dd55-4fcc-88d6-1efea11d13db</td>\n",
       "      <td>808, smooth, r&amp;b, bass, soul, blues, male voca...</td>\n",
       "      <td>(Verse)\\nDancing so close, I can feel your bod...</td>\n",
       "      <td>r&amp;b</td>\n",
       "      <td>False</td>\n",
       "      <td>[0.23394823, 0.01314763, -0.028303534, 0.08800...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4e7aaa68-8bd1-48e1-8f5f-585214a202bf</td>\n",
       "      <td>808, smooth, r&amp;b, bass, soul, blues, male voca...</td>\n",
       "      <td>(Verse)\\nDancing so close, I can feel your bod...</td>\n",
       "      <td>r&amp;b</td>\n",
       "      <td>False</td>\n",
       "      <td>[0.10862688, 0.07702735, -0.040942892, 0.04634...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0219984d-4e98-4279-b549-ce8cbad00cb5</td>\n",
       "      <td>808, smooth, r&amp;b, bass, soul, blues, male voca...</td>\n",
       "      <td>(Verse)\\nDancing so close, I can feel your bod...</td>\n",
       "      <td>r&amp;b</td>\n",
       "      <td>False</td>\n",
       "      <td>[0.12649839, -0.0065048095, 0.0054703723, 0.01...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>70bc6079-7e32-48f2-ab8b-fba02669fe9a</td>\n",
       "      <td>808, smooth, r&amp;b, bass, soul, blues, male voca...</td>\n",
       "      <td>(Verse)\\nDancing so close, I can feel your bod...</td>\n",
       "      <td>r&amp;b</td>\n",
       "      <td>False</td>\n",
       "      <td>[0.13511904, 0.0890281, 0.00094997126, 0.02383...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>74326bb6-56bb-4b47-98cd-03f74091b85d</td>\n",
       "      <td>808, smooth, r&amp;b, bass, soul, blues, male voca...</td>\n",
       "      <td>(Verse)\\nDancing so close, I can feel your bod...</td>\n",
       "      <td>r&amp;b</td>\n",
       "      <td>False</td>\n",
       "      <td>[0.16042964, 0.003049712, -0.091514714, -0.027...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                  s3_id  \\\n",
       "0  1252d952-dd55-4fcc-88d6-1efea11d13db   \n",
       "1  4e7aaa68-8bd1-48e1-8f5f-585214a202bf   \n",
       "2  0219984d-4e98-4279-b549-ce8cbad00cb5   \n",
       "3  70bc6079-7e32-48f2-ab8b-fba02669fe9a   \n",
       "4  74326bb6-56bb-4b47-98cd-03f74091b85d   \n",
       "\n",
       "                                                tags  \\\n",
       "0  808, smooth, r&b, bass, soul, blues, male voca...   \n",
       "1  808, smooth, r&b, bass, soul, blues, male voca...   \n",
       "2  808, smooth, r&b, bass, soul, blues, male voca...   \n",
       "3  808, smooth, r&b, bass, soul, blues, male voca...   \n",
       "4  808, smooth, r&b, bass, soul, blues, male voca...   \n",
       "\n",
       "                                              lyrics genre  instrumental  \\\n",
       "0  (Verse)\\nDancing so close, I can feel your bod...   r&b         False   \n",
       "1  (Verse)\\nDancing so close, I can feel your bod...   r&b         False   \n",
       "2  (Verse)\\nDancing so close, I can feel your bod...   r&b         False   \n",
       "3  (Verse)\\nDancing so close, I can feel your bod...   r&b         False   \n",
       "4  (Verse)\\nDancing so close, I can feel your bod...   r&b         False   \n",
       "\n",
       "                                         ditto_embed  \n",
       "0  [0.23394823, 0.01314763, -0.028303534, 0.08800...  \n",
       "1  [0.10862688, 0.07702735, -0.040942892, 0.04634...  \n",
       "2  [0.12649839, -0.0065048095, 0.0054703723, 0.01...  \n",
       "3  [0.13511904, 0.0890281, 0.00094997126, 0.02383...  \n",
       "4  [0.16042964, 0.003049712, -0.091514714, -0.027...  "
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gen_df[\"ditto_embed\"] = gen_df.progress_apply(get_task_score, axis=1)\n",
    "gen_df = gen_df[gen_df[\"ditto_embed\"].notna()]\n",
    "gen_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "gen_df = gen_df[gen_df[\"ditto_embed\"].notna()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "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>tags</th>\n",
       "      <th>similarity</th>\n",
       "      <th>instrumental</th>\n",
       "      <th>genre</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>bent wonky french house, relax glitch fx, cho...</td>\n",
       "      <td>0.815747</td>\n",
       "      <td>True</td>\n",
       "      <td>edm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>bent wonky french house, relax glitch fx, cho...</td>\n",
       "      <td>0.636331</td>\n",
       "      <td>True</td>\n",
       "      <td>edm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>bent wonky french house, relax glitch fx, cho...</td>\n",
       "      <td>0.707565</td>\n",
       "      <td>True</td>\n",
       "      <td>edm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>bent wonky french house, relax glitch fx, cho...</td>\n",
       "      <td>0.724932</td>\n",
       "      <td>True</td>\n",
       "      <td>edm</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>bent wonky french house, relax glitch fx, cho...</td>\n",
       "      <td>0.661931</td>\n",
       "      <td>True</td>\n",
       "      <td>edm</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                tags  similarity  \\\n",
       "0   bent wonky french house, relax glitch fx, cho...    0.815747   \n",
       "1   bent wonky french house, relax glitch fx, cho...    0.636331   \n",
       "2   bent wonky french house, relax glitch fx, cho...    0.707565   \n",
       "3   bent wonky french house, relax glitch fx, cho...    0.724932   \n",
       "4   bent wonky french house, relax glitch fx, cho...    0.661931   \n",
       "\n",
       "   instrumental genre  \n",
       "0          True   edm  \n",
       "1          True   edm  \n",
       "2          True   edm  \n",
       "3          True   edm  \n",
       "4          True   edm  "
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gens_by_tag = gen_df.groupby(\"tags\")\n",
    "result_pairs = []\n",
    "for tag, group in gens_by_tag:\n",
    "    # Get all pairwise combinations of rows within this group\n",
    "    pairs = list(combinations(group.itertuples(index=False), 2))\n",
    "    for row_a, row_b in pairs:\n",
    "        self_sim = cosine_similarity(row_a.ditto_embed, row_b.ditto_embed)\n",
    "        result_pairs.append(\n",
    "            {\n",
    "                \"tags\": tag,\n",
    "                \"similarity\": self_sim,\n",
    "                \"instrumental\": row_a.instrumental,\n",
    "                \"genre\": row_a.genre,\n",
    "            }\n",
    "        )\n",
    "pairwise_sim_df = pd.DataFrame(result_pairs)\n",
    "pairwise_sim_df.head()"
   ]
  },
  {
   "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>tags</th>\n",
       "      <th>genre</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>max</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>bent wonky french house, relax glitch fx, cho...</td>\n",
       "      <td>edm</td>\n",
       "      <td>0.661930</td>\n",
       "      <td>0.092122</td>\n",
       "      <td>0.416097</td>\n",
       "      <td>0.887543</td>\n",
       "      <td>190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2000's R&amp;B, Contemporary R&amp;B,  Pop</td>\n",
       "      <td>r&amp;b</td>\n",
       "      <td>0.761562</td>\n",
       "      <td>0.099122</td>\n",
       "      <td>0.489587</td>\n",
       "      <td>0.945842</td>\n",
       "      <td>190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>20th century classical guitar, spanish guitar</td>\n",
       "      <td>classical</td>\n",
       "      <td>0.758446</td>\n",
       "      <td>0.076768</td>\n",
       "      <td>0.494066</td>\n",
       "      <td>0.902562</td>\n",
       "      <td>190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>808, smooth, r&amp;b, bass, soul, blues, male voca...</td>\n",
       "      <td>r&amp;b</td>\n",
       "      <td>0.672691</td>\n",
       "      <td>0.085084</td>\n",
       "      <td>0.415154</td>\n",
       "      <td>0.891510</td>\n",
       "      <td>190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>90s, saxophone, rap</td>\n",
       "      <td>hip hop</td>\n",
       "      <td>0.708886</td>\n",
       "      <td>0.089078</td>\n",
       "      <td>0.427973</td>\n",
       "      <td>0.888069</td>\n",
       "      <td>190</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                tags      genre      mean  \\\n",
       "0   bent wonky french house, relax glitch fx, cho...        edm  0.661930   \n",
       "1                 2000's R&B, Contemporary R&B,  Pop        r&b  0.761562   \n",
       "2      20th century classical guitar, spanish guitar  classical  0.758446   \n",
       "3  808, smooth, r&b, bass, soul, blues, male voca...        r&b  0.672691   \n",
       "4                                90s, saxophone, rap    hip hop  0.708886   \n",
       "\n",
       "        std       min       max  count  \n",
       "0  0.092122  0.416097  0.887543    190  \n",
       "1  0.099122  0.489587  0.945842    190  \n",
       "2  0.076768  0.494066  0.902562    190  \n",
       "3  0.085084  0.415154  0.891510    190  \n",
       "4  0.089078  0.427973  0.888069    190  "
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_by_tag = pairwise_sim_df.groupby([\"tags\", \"genre\"])\n",
    "stats_by_tag = results_by_tag[\"similarity\"].agg([\"mean\", \"std\", \"min\", \"max\", \"count\"]).reset_index()\n",
    "stats_by_tag.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(15, 6))\n",
    "\n",
    "# Step 3: Scatter plot for each genre, using mean similarity and size by std_dev\n",
    "for genre in stats_by_tag[\"genre\"].unique():\n",
    "    genre_data = stats_by_tag[stats_by_tag[\"genre\"] == genre]\n",
    "    plt.scatter(\n",
    "        genre_data[\"genre\"],  # X-axis: genre\n",
    "        genre_data[\"mean\"],  # Y-axis: mean similarity\n",
    "        s=genre_data[\"std\"] ** 2 * 3.14159 * 3000,  # Circle size: scaled standard deviation\n",
    "        alpha=0.5,  # Transparency\n",
    "        label=genre,  # Label for each genre\n",
    "    )\n",
    "\n",
    "# Step 4: Customize the plot\n",
    "plt.xlabel(\"Genre\")\n",
    "plt.ylabel(\"Mean Cosine Similarity\")\n",
    "plt.title(f\"Pairwise Ditto self_sim of Tags Grouped by Genre: {MODEL}\")\n",
    "# plt.legend(title='Genre')\n",
    "plt.ylim([0.0, 1.0])\n",
    "plt.grid(True)\n",
    "\n",
    "# Show the plot\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "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>genre</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>max</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>ambient</td>\n",
       "      <td>0.661179</td>\n",
       "      <td>0.147021</td>\n",
       "      <td>0.152099</td>\n",
       "      <td>0.931328</td>\n",
       "      <td>950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>classical</td>\n",
       "      <td>0.668661</td>\n",
       "      <td>0.113829</td>\n",
       "      <td>0.319572</td>\n",
       "      <td>0.942904</td>\n",
       "      <td>950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>country</td>\n",
       "      <td>0.647191</td>\n",
       "      <td>0.122487</td>\n",
       "      <td>0.236680</td>\n",
       "      <td>0.940527</td>\n",
       "      <td>950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>dance</td>\n",
       "      <td>0.787895</td>\n",
       "      <td>0.104898</td>\n",
       "      <td>0.394291</td>\n",
       "      <td>0.967613</td>\n",
       "      <td>950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>edm</td>\n",
       "      <td>0.757610</td>\n",
       "      <td>0.113427</td>\n",
       "      <td>0.293308</td>\n",
       "      <td>0.952807</td>\n",
       "      <td>950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>electronic</td>\n",
       "      <td>0.826205</td>\n",
       "      <td>0.095285</td>\n",
       "      <td>0.513479</td>\n",
       "      <td>0.963025</td>\n",
       "      <td>950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>folk</td>\n",
       "      <td>0.653398</td>\n",
       "      <td>0.135954</td>\n",
       "      <td>0.208859</td>\n",
       "      <td>0.942047</td>\n",
       "      <td>950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>hip hop</td>\n",
       "      <td>0.682946</td>\n",
       "      <td>0.132669</td>\n",
       "      <td>0.236874</td>\n",
       "      <td>0.922127</td>\n",
       "      <td>950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>jazz</td>\n",
       "      <td>0.639149</td>\n",
       "      <td>0.108810</td>\n",
       "      <td>0.266914</td>\n",
       "      <td>0.913913</td>\n",
       "      <td>950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>metal</td>\n",
       "      <td>0.663300</td>\n",
       "      <td>0.103600</td>\n",
       "      <td>0.390698</td>\n",
       "      <td>0.906260</td>\n",
       "      <td>950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>pop</td>\n",
       "      <td>0.637877</td>\n",
       "      <td>0.136786</td>\n",
       "      <td>0.191770</td>\n",
       "      <td>0.946130</td>\n",
       "      <td>950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>punk</td>\n",
       "      <td>0.679919</td>\n",
       "      <td>0.092671</td>\n",
       "      <td>0.324275</td>\n",
       "      <td>0.904922</td>\n",
       "      <td>950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>r&amp;b</td>\n",
       "      <td>0.703166</td>\n",
       "      <td>0.097646</td>\n",
       "      <td>0.415154</td>\n",
       "      <td>0.945842</td>\n",
       "      <td>950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>reggae</td>\n",
       "      <td>0.642362</td>\n",
       "      <td>0.113826</td>\n",
       "      <td>0.244167</td>\n",
       "      <td>0.898173</td>\n",
       "      <td>950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>rock</td>\n",
       "      <td>0.654479</td>\n",
       "      <td>0.108416</td>\n",
       "      <td>0.258792</td>\n",
       "      <td>0.938286</td>\n",
       "      <td>950</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         genre      mean       std       min       max  count\n",
       "0      ambient  0.661179  0.147021  0.152099  0.931328    950\n",
       "1    classical  0.668661  0.113829  0.319572  0.942904    950\n",
       "2      country  0.647191  0.122487  0.236680  0.940527    950\n",
       "3        dance  0.787895  0.104898  0.394291  0.967613    950\n",
       "4          edm  0.757610  0.113427  0.293308  0.952807    950\n",
       "5   electronic  0.826205  0.095285  0.513479  0.963025    950\n",
       "6         folk  0.653398  0.135954  0.208859  0.942047    950\n",
       "7      hip hop  0.682946  0.132669  0.236874  0.922127    950\n",
       "8         jazz  0.639149  0.108810  0.266914  0.913913    950\n",
       "9        metal  0.663300  0.103600  0.390698  0.906260    950\n",
       "10         pop  0.637877  0.136786  0.191770  0.946130    950\n",
       "11        punk  0.679919  0.092671  0.324275  0.904922    950\n",
       "12         r&b  0.703166  0.097646  0.415154  0.945842    950\n",
       "13      reggae  0.642362  0.113826  0.244167  0.898173    950\n",
       "14        rock  0.654479  0.108416  0.258792  0.938286    950"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_by_genre = pairwise_sim_df.groupby([\"genre\"])\n",
    "stats_by_genre = results_by_genre[\"similarity\"].agg([\"mean\", \"std\", \"min\", \"max\", \"count\"]).reset_index()\n",
    "stats_by_genre.head(15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1500x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(15, 6))\n",
    "plt.bar(stats_by_genre[\"genre\"], stats_by_genre[\"mean\"], color=\"skyblue\")\n",
    "\n",
    "plt.xlabel(\"Genre\")\n",
    "plt.ylabel(\"Mean Pairwise Similarity\")\n",
    "plt.title(f\"Pairwise Ditto self_sim Across Genres: {MODEL}\")\n",
    "plt.xticks(rotation=0)  # Rotate x-axis labels for better readability if needed\n",
    "plt.ylim([0.0, 1.0])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "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>instrumental</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>max</th>\n",
       "      <th>count</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>False</td>\n",
       "      <td>0.688348</td>\n",
       "      <td>0.122911</td>\n",
       "      <td>0.191770</td>\n",
       "      <td>0.967613</td>\n",
       "      <td>8930</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>True</td>\n",
       "      <td>0.684798</td>\n",
       "      <td>0.138486</td>\n",
       "      <td>0.152099</td>\n",
       "      <td>0.963025</td>\n",
       "      <td>5320</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   instrumental      mean       std       min       max  count\n",
       "0         False  0.688348  0.122911  0.191770  0.967613   8930\n",
       "1          True  0.684798  0.138486  0.152099  0.963025   5320"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results_by_lyrics = pairwise_sim_df.groupby([\"instrumental\"])\n",
    "stats_by_lyrics = (\n",
    "    results_by_lyrics[\"similarity\"].agg([\"mean\", \"std\", \"min\", \"max\", \"count\"]).reset_index()\n",
    ")\n",
    "stats_by_lyrics.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "suno_clean",
   "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": "3.10.15"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}