similartiy_ratings.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from itertools import combinations\n",
    "import openai\n",
    "import base64\n",
    "import numpy as np\n",
    "from tenacity import (\n",
    "    retry,\n",
    "    stop_after_attempt,\n",
    "    wait_random_exponential,\n",
    ")\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "human_data = pd.read_csv(\"data\\\\30_rocks_similarities.csv\", index_col=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>I_Andesite_06</th>\n",
       "      <th>I_Basalt_02</th>\n",
       "      <th>I_Diorite_10</th>\n",
       "      <th>I_Gabbro_03</th>\n",
       "      <th>I_Granite_11</th>\n",
       "      <th>I_Obsidian_09</th>\n",
       "      <th>I_Pegmatite_01</th>\n",
       "      <th>I_Peridotite_11</th>\n",
       "      <th>I_Pumice_06</th>\n",
       "      <th>I_Rhyolite_11</th>\n",
       "      <th>...</th>\n",
       "      <th>S_Bituminous Coal_10</th>\n",
       "      <th>S_Breccia_11</th>\n",
       "      <th>S_Chert_03</th>\n",
       "      <th>S_Conglomerate_08</th>\n",
       "      <th>S_Dolomite_12</th>\n",
       "      <th>S_Micrite_09</th>\n",
       "      <th>S_Rock Gypsum_02</th>\n",
       "      <th>S_Rock Salt_12</th>\n",
       "      <th>S_Sandstone_12</th>\n",
       "      <th>S_Shale_04</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>I_Andesite_06</th>\n",
       "      <td>NaN</td>\n",
       "      <td>5.5294</td>\n",
       "      <td>6.7059</td>\n",
       "      <td>6.8824</td>\n",
       "      <td>5.6324</td>\n",
       "      <td>3.7206</td>\n",
       "      <td>3.2500</td>\n",
       "      <td>5.1765</td>\n",
       "      <td>3.1912</td>\n",
       "      <td>5.5441</td>\n",
       "      <td>...</td>\n",
       "      <td>4.0294</td>\n",
       "      <td>4.2059</td>\n",
       "      <td>2.7941</td>\n",
       "      <td>3.4118</td>\n",
       "      <td>4.1912</td>\n",
       "      <td>2.8676</td>\n",
       "      <td>3.3676</td>\n",
       "      <td>5.0000</td>\n",
       "      <td>2.4118</td>\n",
       "      <td>4.7206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>I_Basalt_02</th>\n",
       "      <td>5.5294</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.7353</td>\n",
       "      <td>5.8382</td>\n",
       "      <td>2.9559</td>\n",
       "      <td>5.3676</td>\n",
       "      <td>2.8382</td>\n",
       "      <td>3.9706</td>\n",
       "      <td>3.1765</td>\n",
       "      <td>6.5441</td>\n",
       "      <td>...</td>\n",
       "      <td>5.7353</td>\n",
       "      <td>2.6765</td>\n",
       "      <td>3.1324</td>\n",
       "      <td>2.6029</td>\n",
       "      <td>5.7500</td>\n",
       "      <td>6.1618</td>\n",
       "      <td>4.6765</td>\n",
       "      <td>3.8824</td>\n",
       "      <td>3.2941</td>\n",
       "      <td>6.7353</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>I_Diorite_10</th>\n",
       "      <td>6.7059</td>\n",
       "      <td>3.7353</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.8382</td>\n",
       "      <td>6.5735</td>\n",
       "      <td>3.1176</td>\n",
       "      <td>4.4559</td>\n",
       "      <td>5.3529</td>\n",
       "      <td>3.8529</td>\n",
       "      <td>3.5735</td>\n",
       "      <td>...</td>\n",
       "      <td>3.7206</td>\n",
       "      <td>3.2353</td>\n",
       "      <td>3.2353</td>\n",
       "      <td>4.5000</td>\n",
       "      <td>4.6176</td>\n",
       "      <td>3.7353</td>\n",
       "      <td>3.2647</td>\n",
       "      <td>3.7500</td>\n",
       "      <td>2.6471</td>\n",
       "      <td>3.7059</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>I_Gabbro_03</th>\n",
       "      <td>6.8824</td>\n",
       "      <td>5.8382</td>\n",
       "      <td>6.8382</td>\n",
       "      <td>NaN</td>\n",
       "      <td>6.0882</td>\n",
       "      <td>3.8971</td>\n",
       "      <td>4.0294</td>\n",
       "      <td>4.7206</td>\n",
       "      <td>3.3971</td>\n",
       "      <td>4.0735</td>\n",
       "      <td>...</td>\n",
       "      <td>4.9559</td>\n",
       "      <td>4.0588</td>\n",
       "      <td>2.8382</td>\n",
       "      <td>3.3676</td>\n",
       "      <td>4.1324</td>\n",
       "      <td>3.6176</td>\n",
       "      <td>3.7794</td>\n",
       "      <td>3.2206</td>\n",
       "      <td>2.2794</td>\n",
       "      <td>5.0294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>I_Granite_11</th>\n",
       "      <td>5.6324</td>\n",
       "      <td>2.9559</td>\n",
       "      <td>6.5735</td>\n",
       "      <td>6.0882</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.5147</td>\n",
       "      <td>4.7059</td>\n",
       "      <td>3.6618</td>\n",
       "      <td>3.5000</td>\n",
       "      <td>3.4706</td>\n",
       "      <td>...</td>\n",
       "      <td>3.2353</td>\n",
       "      <td>4.3235</td>\n",
       "      <td>3.0000</td>\n",
       "      <td>3.9853</td>\n",
       "      <td>4.1618</td>\n",
       "      <td>3.1618</td>\n",
       "      <td>4.3676</td>\n",
       "      <td>4.3235</td>\n",
       "      <td>2.5882</td>\n",
       "      <td>2.7500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>I_Obsidian_09</th>\n",
       "      <td>3.7206</td>\n",
       "      <td>5.3676</td>\n",
       "      <td>3.1176</td>\n",
       "      <td>3.8971</td>\n",
       "      <td>2.5147</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.2059</td>\n",
       "      <td>3.2059</td>\n",
       "      <td>2.3676</td>\n",
       "      <td>3.9265</td>\n",
       "      <td>...</td>\n",
       "      <td>7.0588</td>\n",
       "      <td>2.3235</td>\n",
       "      <td>4.1912</td>\n",
       "      <td>2.9559</td>\n",
       "      <td>3.2059</td>\n",
       "      <td>4.4118</td>\n",
       "      <td>4.0441</td>\n",
       "      <td>3.7794</td>\n",
       "      <td>2.4706</td>\n",
       "      <td>5.6029</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>I_Pegmatite_01</th>\n",
       "      <td>3.2500</td>\n",
       "      <td>2.8382</td>\n",
       "      <td>4.4559</td>\n",
       "      <td>4.0294</td>\n",
       "      <td>4.7059</td>\n",
       "      <td>3.2059</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.7059</td>\n",
       "      <td>3.8382</td>\n",
       "      <td>2.7059</td>\n",
       "      <td>...</td>\n",
       "      <td>3.2353</td>\n",
       "      <td>2.8971</td>\n",
       "      <td>4.2941</td>\n",
       "      <td>5.8676</td>\n",
       "      <td>2.7941</td>\n",
       "      <td>3.5000</td>\n",
       "      <td>2.9118</td>\n",
       "      <td>3.3971</td>\n",
       "      <td>2.7059</td>\n",
       "      <td>2.7059</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>I_Peridotite_11</th>\n",
       "      <td>5.1765</td>\n",
       "      <td>3.9706</td>\n",
       "      <td>5.3529</td>\n",
       "      <td>4.7206</td>\n",
       "      <td>3.6618</td>\n",
       "      <td>3.2059</td>\n",
       "      <td>5.7059</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.5735</td>\n",
       "      <td>3.1765</td>\n",
       "      <td>...</td>\n",
       "      <td>3.4265</td>\n",
       "      <td>2.1765</td>\n",
       "      <td>3.3382</td>\n",
       "      <td>6.1471</td>\n",
       "      <td>3.3676</td>\n",
       "      <td>4.1618</td>\n",
       "      <td>2.7647</td>\n",
       "      <td>2.5294</td>\n",
       "      <td>2.9559</td>\n",
       "      <td>3.2941</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>I_Pumice_06</th>\n",
       "      <td>3.1912</td>\n",
       "      <td>3.1765</td>\n",
       "      <td>3.8529</td>\n",
       "      <td>3.3971</td>\n",
       "      <td>3.5000</td>\n",
       "      <td>2.3676</td>\n",
       "      <td>3.8382</td>\n",
       "      <td>5.5735</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.8382</td>\n",
       "      <td>...</td>\n",
       "      <td>2.9706</td>\n",
       "      <td>2.1029</td>\n",
       "      <td>4.3088</td>\n",
       "      <td>4.5441</td>\n",
       "      <td>4.3382</td>\n",
       "      <td>4.2500</td>\n",
       "      <td>2.9706</td>\n",
       "      <td>2.1324</td>\n",
       "      <td>5.5147</td>\n",
       "      <td>2.3382</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>I_Rhyolite_11</th>\n",
       "      <td>5.5441</td>\n",
       "      <td>6.5441</td>\n",
       "      <td>3.5735</td>\n",
       "      <td>4.0735</td>\n",
       "      <td>3.4706</td>\n",
       "      <td>3.9265</td>\n",
       "      <td>2.7059</td>\n",
       "      <td>3.1765</td>\n",
       "      <td>3.8382</td>\n",
       "      <td>NaN</td>\n",
       "      <td>...</td>\n",
       "      <td>4.8382</td>\n",
       "      <td>2.3235</td>\n",
       "      <td>3.8529</td>\n",
       "      <td>2.9559</td>\n",
       "      <td>6.9853</td>\n",
       "      <td>4.4559</td>\n",
       "      <td>4.9853</td>\n",
       "      <td>3.7794</td>\n",
       "      <td>3.9118</td>\n",
       "      <td>5.5294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>M_Amphibolite_07</th>\n",
       "      <td>7.4118</td>\n",
       "      <td>5.5735</td>\n",
       "      <td>6.3382</td>\n",
       "      <td>7.2941</td>\n",
       "      <td>5.6029</td>\n",
       "      <td>4.7353</td>\n",
       "      <td>4.0441</td>\n",
       "      <td>4.8824</td>\n",
       "      <td>3.0882</td>\n",
       "      <td>4.2647</td>\n",
       "      <td>...</td>\n",
       "      <td>5.0588</td>\n",
       "      <td>3.8235</td>\n",
       "      <td>2.9853</td>\n",
       "      <td>3.4412</td>\n",
       "      <td>3.1765</td>\n",
       "      <td>2.9265</td>\n",
       "      <td>3.3971</td>\n",
       "      <td>4.6765</td>\n",
       "      <td>2.3382</td>\n",
       "      <td>4.9412</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>M_Anthracite_12</th>\n",
       "      <td>4.0735</td>\n",
       "      <td>5.4706</td>\n",
       "      <td>3.4265</td>\n",
       "      <td>4.5441</td>\n",
       "      <td>3.0735</td>\n",
       "      <td>8.0000</td>\n",
       "      <td>3.3235</td>\n",
       "      <td>2.9265</td>\n",
       "      <td>2.3971</td>\n",
       "      <td>3.7059</td>\n",
       "      <td>...</td>\n",
       "      <td>6.8824</td>\n",
       "      <td>2.4412</td>\n",
       "      <td>3.0147</td>\n",
       "      <td>2.6618</td>\n",
       "      <td>2.9118</td>\n",
       "      <td>3.9118</td>\n",
       "      <td>3.5588</td>\n",
       "      <td>4.5000</td>\n",
       "      <td>2.1324</td>\n",
       "      <td>5.2500</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>M_Gneiss_07</th>\n",
       "      <td>3.6912</td>\n",
       "      <td>4.5882</td>\n",
       "      <td>4.8529</td>\n",
       "      <td>5.6176</td>\n",
       "      <td>3.8235</td>\n",
       "      <td>4.0735</td>\n",
       "      <td>4.6471</td>\n",
       "      <td>4.5147</td>\n",
       "      <td>4.5000</td>\n",
       "      <td>4.5588</td>\n",
       "      <td>...</td>\n",
       "      <td>4.5735</td>\n",
       "      <td>2.7059</td>\n",
       "      <td>3.9706</td>\n",
       "      <td>4.3382</td>\n",
       "      <td>4.4559</td>\n",
       "      <td>5.3824</td>\n",
       "      <td>3.2794</td>\n",
       "      <td>2.5735</td>\n",
       "      <td>3.2794</td>\n",
       "      <td>4.6324</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>M_Hornfels_08</th>\n",
       "      <td>5.6324</td>\n",
       "      <td>7.0294</td>\n",
       "      <td>4.1324</td>\n",
       "      <td>5.7206</td>\n",
       "      <td>3.7206</td>\n",
       "      <td>5.5147</td>\n",
       "      <td>3.9559</td>\n",
       "      <td>4.4559</td>\n",
       "      <td>3.0882</td>\n",
       "      <td>5.8824</td>\n",
       "      <td>...</td>\n",
       "      <td>6.8382</td>\n",
       "      <td>3.1618</td>\n",
       "      <td>3.1176</td>\n",
       "      <td>3.6765</td>\n",
       "      <td>4.7941</td>\n",
       "      <td>5.5000</td>\n",
       "      <td>3.9706</td>\n",
       "      <td>3.6618</td>\n",
       "      <td>2.8088</td>\n",
       "      <td>5.3971</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>M_Marble_04</th>\n",
       "      <td>4.4118</td>\n",
       "      <td>4.9853</td>\n",
       "      <td>3.3235</td>\n",
       "      <td>4.5147</td>\n",
       "      <td>5.1765</td>\n",
       "      <td>3.3382</td>\n",
       "      <td>2.5588</td>\n",
       "      <td>2.6471</td>\n",
       "      <td>2.9412</td>\n",
       "      <td>5.5000</td>\n",
       "      <td>...</td>\n",
       "      <td>3.0882</td>\n",
       "      <td>3.7794</td>\n",
       "      <td>3.1618</td>\n",
       "      <td>2.4559</td>\n",
       "      <td>5.5441</td>\n",
       "      <td>3.9412</td>\n",
       "      <td>7.0882</td>\n",
       "      <td>6.4706</td>\n",
       "      <td>2.6176</td>\n",
       "      <td>2.9118</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>M_Migmatite_08</th>\n",
       "      <td>5.2500</td>\n",
       "      <td>5.0294</td>\n",
       "      <td>5.7941</td>\n",
       "      <td>6.1324</td>\n",
       "      <td>4.8676</td>\n",
       "      <td>4.0000</td>\n",
       "      <td>4.8235</td>\n",
       "      <td>5.5882</td>\n",
       "      <td>4.1176</td>\n",
       "      <td>4.3088</td>\n",
       "      <td>...</td>\n",
       "      <td>5.0294</td>\n",
       "      <td>2.2941</td>\n",
       "      <td>3.5735</td>\n",
       "      <td>4.2206</td>\n",
       "      <td>5.1618</td>\n",
       "      <td>4.4853</td>\n",
       "      <td>3.0735</td>\n",
       "      <td>2.9265</td>\n",
       "      <td>3.5588</td>\n",
       "      <td>4.4559</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>M_Phyllite_09</th>\n",
       "      <td>4.2059</td>\n",
       "      <td>5.0441</td>\n",
       "      <td>5.6912</td>\n",
       "      <td>4.7941</td>\n",
       "      <td>3.1912</td>\n",
       "      <td>4.3382</td>\n",
       "      <td>4.3382</td>\n",
       "      <td>6.7206</td>\n",
       "      <td>4.5588</td>\n",
       "      <td>4.5000</td>\n",
       "      <td>...</td>\n",
       "      <td>4.6765</td>\n",
       "      <td>2.1765</td>\n",
       "      <td>3.8529</td>\n",
       "      <td>4.5147</td>\n",
       "      <td>4.3676</td>\n",
       "      <td>5.4559</td>\n",
       "      <td>3.0882</td>\n",
       "      <td>2.9265</td>\n",
       "      <td>3.9706</td>\n",
       "      <td>3.8088</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>M_Quartzite_09</th>\n",
       "      <td>6.4118</td>\n",
       "      <td>5.9706</td>\n",
       "      <td>4.9559</td>\n",
       "      <td>6.2647</td>\n",
       "      <td>4.9265</td>\n",
       "      <td>3.3235</td>\n",
       "      <td>3.1618</td>\n",
       "      <td>3.4118</td>\n",
       "      <td>2.9706</td>\n",
       "      <td>5.8382</td>\n",
       "      <td>...</td>\n",
       "      <td>3.3529</td>\n",
       "      <td>4.1912</td>\n",
       "      <td>3.0294</td>\n",
       "      <td>2.9412</td>\n",
       "      <td>5.1912</td>\n",
       "      <td>3.8382</td>\n",
       "      <td>5.5882</td>\n",
       "      <td>5.0147</td>\n",
       "      <td>2.6176</td>\n",
       "      <td>4.0294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>M_Schist_02</th>\n",
       "      <td>6.7206</td>\n",
       "      <td>5.0000</td>\n",
       "      <td>5.8676</td>\n",
       "      <td>6.1471</td>\n",
       "      <td>5.0735</td>\n",
       "      <td>4.0294</td>\n",
       "      <td>5.4559</td>\n",
       "      <td>5.5441</td>\n",
       "      <td>3.6618</td>\n",
       "      <td>5.4265</td>\n",
       "      <td>...</td>\n",
       "      <td>4.7059</td>\n",
       "      <td>3.1912</td>\n",
       "      <td>3.3235</td>\n",
       "      <td>5.2206</td>\n",
       "      <td>4.8529</td>\n",
       "      <td>3.7206</td>\n",
       "      <td>3.8529</td>\n",
       "      <td>5.0882</td>\n",
       "      <td>2.6324</td>\n",
       "      <td>3.9412</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>M_Slate_09</th>\n",
       "      <td>4.5147</td>\n",
       "      <td>5.4265</td>\n",
       "      <td>3.1765</td>\n",
       "      <td>3.6618</td>\n",
       "      <td>2.8824</td>\n",
       "      <td>3.2500</td>\n",
       "      <td>2.8824</td>\n",
       "      <td>2.8088</td>\n",
       "      <td>2.9412</td>\n",
       "      <td>6.5147</td>\n",
       "      <td>...</td>\n",
       "      <td>3.8676</td>\n",
       "      <td>2.0588</td>\n",
       "      <td>3.7941</td>\n",
       "      <td>2.7941</td>\n",
       "      <td>5.4706</td>\n",
       "      <td>4.2059</td>\n",
       "      <td>4.1029</td>\n",
       "      <td>3.5588</td>\n",
       "      <td>2.7059</td>\n",
       "      <td>5.8088</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S_Bituminous Coal_10</th>\n",
       "      <td>4.0294</td>\n",
       "      <td>5.7353</td>\n",
       "      <td>3.7206</td>\n",
       "      <td>4.9559</td>\n",
       "      <td>3.2353</td>\n",
       "      <td>7.0588</td>\n",
       "      <td>3.2353</td>\n",
       "      <td>3.4265</td>\n",
       "      <td>2.9706</td>\n",
       "      <td>4.8382</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.2647</td>\n",
       "      <td>3.0882</td>\n",
       "      <td>2.9706</td>\n",
       "      <td>4.3971</td>\n",
       "      <td>4.7941</td>\n",
       "      <td>3.0147</td>\n",
       "      <td>3.2353</td>\n",
       "      <td>3.0441</td>\n",
       "      <td>6.8382</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S_Breccia_11</th>\n",
       "      <td>4.2059</td>\n",
       "      <td>2.6765</td>\n",
       "      <td>3.2353</td>\n",
       "      <td>4.0588</td>\n",
       "      <td>4.3235</td>\n",
       "      <td>2.3235</td>\n",
       "      <td>2.8971</td>\n",
       "      <td>2.1765</td>\n",
       "      <td>2.1029</td>\n",
       "      <td>2.3235</td>\n",
       "      <td>...</td>\n",
       "      <td>2.2647</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.1912</td>\n",
       "      <td>2.8382</td>\n",
       "      <td>2.5147</td>\n",
       "      <td>2.7353</td>\n",
       "      <td>3.2206</td>\n",
       "      <td>2.6765</td>\n",
       "      <td>2.1471</td>\n",
       "      <td>2.1618</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S_Chert_03</th>\n",
       "      <td>2.7941</td>\n",
       "      <td>3.1324</td>\n",
       "      <td>3.2353</td>\n",
       "      <td>2.8382</td>\n",
       "      <td>3.0000</td>\n",
       "      <td>4.1912</td>\n",
       "      <td>4.2941</td>\n",
       "      <td>3.3382</td>\n",
       "      <td>4.3088</td>\n",
       "      <td>3.8529</td>\n",
       "      <td>...</td>\n",
       "      <td>3.0882</td>\n",
       "      <td>2.1912</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.8088</td>\n",
       "      <td>4.3235</td>\n",
       "      <td>5.7647</td>\n",
       "      <td>4.4853</td>\n",
       "      <td>2.9265</td>\n",
       "      <td>4.6618</td>\n",
       "      <td>3.2794</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S_Conglomerate_08</th>\n",
       "      <td>3.4118</td>\n",
       "      <td>2.6029</td>\n",
       "      <td>4.5000</td>\n",
       "      <td>3.3676</td>\n",
       "      <td>3.9853</td>\n",
       "      <td>2.9559</td>\n",
       "      <td>5.8676</td>\n",
       "      <td>6.1471</td>\n",
       "      <td>4.5441</td>\n",
       "      <td>2.9559</td>\n",
       "      <td>...</td>\n",
       "      <td>2.9706</td>\n",
       "      <td>2.8382</td>\n",
       "      <td>3.8088</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.8088</td>\n",
       "      <td>4.4265</td>\n",
       "      <td>2.1912</td>\n",
       "      <td>2.6765</td>\n",
       "      <td>3.1029</td>\n",
       "      <td>2.6471</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S_Dolomite_12</th>\n",
       "      <td>4.1912</td>\n",
       "      <td>5.7500</td>\n",
       "      <td>4.6176</td>\n",
       "      <td>4.1324</td>\n",
       "      <td>4.1618</td>\n",
       "      <td>3.2059</td>\n",
       "      <td>2.7941</td>\n",
       "      <td>3.3676</td>\n",
       "      <td>4.3382</td>\n",
       "      <td>6.9853</td>\n",
       "      <td>...</td>\n",
       "      <td>4.3971</td>\n",
       "      <td>2.5147</td>\n",
       "      <td>4.3235</td>\n",
       "      <td>2.8088</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.7059</td>\n",
       "      <td>5.3235</td>\n",
       "      <td>3.7794</td>\n",
       "      <td>3.9412</td>\n",
       "      <td>5.5882</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S_Micrite_09</th>\n",
       "      <td>2.8676</td>\n",
       "      <td>6.1618</td>\n",
       "      <td>3.7353</td>\n",
       "      <td>3.6176</td>\n",
       "      <td>3.1618</td>\n",
       "      <td>4.4118</td>\n",
       "      <td>3.5000</td>\n",
       "      <td>4.1618</td>\n",
       "      <td>4.2500</td>\n",
       "      <td>4.4559</td>\n",
       "      <td>...</td>\n",
       "      <td>4.7941</td>\n",
       "      <td>2.7353</td>\n",
       "      <td>5.7647</td>\n",
       "      <td>4.4265</td>\n",
       "      <td>5.7059</td>\n",
       "      <td>NaN</td>\n",
       "      <td>4.3235</td>\n",
       "      <td>2.5882</td>\n",
       "      <td>4.0588</td>\n",
       "      <td>5.1765</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S_Rock Gypsum_02</th>\n",
       "      <td>3.3676</td>\n",
       "      <td>4.6765</td>\n",
       "      <td>3.2647</td>\n",
       "      <td>3.7794</td>\n",
       "      <td>4.3676</td>\n",
       "      <td>4.0441</td>\n",
       "      <td>2.9118</td>\n",
       "      <td>2.7647</td>\n",
       "      <td>2.9706</td>\n",
       "      <td>4.9853</td>\n",
       "      <td>...</td>\n",
       "      <td>3.0147</td>\n",
       "      <td>3.2206</td>\n",
       "      <td>4.4853</td>\n",
       "      <td>2.1912</td>\n",
       "      <td>5.3235</td>\n",
       "      <td>4.3235</td>\n",
       "      <td>NaN</td>\n",
       "      <td>5.7794</td>\n",
       "      <td>2.5294</td>\n",
       "      <td>3.1029</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S_Rock Salt_12</th>\n",
       "      <td>5.0000</td>\n",
       "      <td>3.8824</td>\n",
       "      <td>3.7500</td>\n",
       "      <td>3.2206</td>\n",
       "      <td>4.3235</td>\n",
       "      <td>3.7794</td>\n",
       "      <td>3.3971</td>\n",
       "      <td>2.5294</td>\n",
       "      <td>2.1324</td>\n",
       "      <td>3.7794</td>\n",
       "      <td>...</td>\n",
       "      <td>3.2353</td>\n",
       "      <td>2.6765</td>\n",
       "      <td>2.9265</td>\n",
       "      <td>2.6765</td>\n",
       "      <td>3.7794</td>\n",
       "      <td>2.5882</td>\n",
       "      <td>5.7794</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.8676</td>\n",
       "      <td>2.9412</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S_Sandstone_12</th>\n",
       "      <td>2.4118</td>\n",
       "      <td>3.2941</td>\n",
       "      <td>2.6471</td>\n",
       "      <td>2.2794</td>\n",
       "      <td>2.5882</td>\n",
       "      <td>2.4706</td>\n",
       "      <td>2.7059</td>\n",
       "      <td>2.9559</td>\n",
       "      <td>5.5147</td>\n",
       "      <td>3.9118</td>\n",
       "      <td>...</td>\n",
       "      <td>3.0441</td>\n",
       "      <td>2.1471</td>\n",
       "      <td>4.6618</td>\n",
       "      <td>3.1029</td>\n",
       "      <td>3.9412</td>\n",
       "      <td>4.0588</td>\n",
       "      <td>2.5294</td>\n",
       "      <td>1.8676</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2.7794</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>S_Shale_04</th>\n",
       "      <td>4.7206</td>\n",
       "      <td>6.7353</td>\n",
       "      <td>3.7059</td>\n",
       "      <td>5.0294</td>\n",
       "      <td>2.7500</td>\n",
       "      <td>5.6029</td>\n",
       "      <td>2.7059</td>\n",
       "      <td>3.2941</td>\n",
       "      <td>2.3382</td>\n",
       "      <td>5.5294</td>\n",
       "      <td>...</td>\n",
       "      <td>6.8382</td>\n",
       "      <td>2.1618</td>\n",
       "      <td>3.2794</td>\n",
       "      <td>2.6471</td>\n",
       "      <td>5.5882</td>\n",
       "      <td>5.1765</td>\n",
       "      <td>3.1029</td>\n",
       "      <td>2.9412</td>\n",
       "      <td>2.7794</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>30 rows × 30 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                      I_Andesite_06  I_Basalt_02  I_Diorite_10  I_Gabbro_03  \\\n",
       "I_Andesite_06                   NaN       5.5294        6.7059       6.8824   \n",
       "I_Basalt_02                  5.5294          NaN        3.7353       5.8382   \n",
       "I_Diorite_10                 6.7059       3.7353           NaN       6.8382   \n",
       "I_Gabbro_03                  6.8824       5.8382        6.8382          NaN   \n",
       "I_Granite_11                 5.6324       2.9559        6.5735       6.0882   \n",
       "I_Obsidian_09                3.7206       5.3676        3.1176       3.8971   \n",
       "I_Pegmatite_01               3.2500       2.8382        4.4559       4.0294   \n",
       "I_Peridotite_11              5.1765       3.9706        5.3529       4.7206   \n",
       "I_Pumice_06                  3.1912       3.1765        3.8529       3.3971   \n",
       "I_Rhyolite_11                5.5441       6.5441        3.5735       4.0735   \n",
       "M_Amphibolite_07             7.4118       5.5735        6.3382       7.2941   \n",
       "M_Anthracite_12              4.0735       5.4706        3.4265       4.5441   \n",
       "M_Gneiss_07                  3.6912       4.5882        4.8529       5.6176   \n",
       "M_Hornfels_08                5.6324       7.0294        4.1324       5.7206   \n",
       "M_Marble_04                  4.4118       4.9853        3.3235       4.5147   \n",
       "M_Migmatite_08               5.2500       5.0294        5.7941       6.1324   \n",
       "M_Phyllite_09                4.2059       5.0441        5.6912       4.7941   \n",
       "M_Quartzite_09               6.4118       5.9706        4.9559       6.2647   \n",
       "M_Schist_02                  6.7206       5.0000        5.8676       6.1471   \n",
       "M_Slate_09                   4.5147       5.4265        3.1765       3.6618   \n",
       "S_Bituminous Coal_10         4.0294       5.7353        3.7206       4.9559   \n",
       "S_Breccia_11                 4.2059       2.6765        3.2353       4.0588   \n",
       "S_Chert_03                   2.7941       3.1324        3.2353       2.8382   \n",
       "S_Conglomerate_08            3.4118       2.6029        4.5000       3.3676   \n",
       "S_Dolomite_12                4.1912       5.7500        4.6176       4.1324   \n",
       "S_Micrite_09                 2.8676       6.1618        3.7353       3.6176   \n",
       "S_Rock Gypsum_02             3.3676       4.6765        3.2647       3.7794   \n",
       "S_Rock Salt_12               5.0000       3.8824        3.7500       3.2206   \n",
       "S_Sandstone_12               2.4118       3.2941        2.6471       2.2794   \n",
       "S_Shale_04                   4.7206       6.7353        3.7059       5.0294   \n",
       "\n",
       "                      I_Granite_11  I_Obsidian_09  I_Pegmatite_01  \\\n",
       "I_Andesite_06               5.6324         3.7206          3.2500   \n",
       "I_Basalt_02                 2.9559         5.3676          2.8382   \n",
       "I_Diorite_10                6.5735         3.1176          4.4559   \n",
       "I_Gabbro_03                 6.0882         3.8971          4.0294   \n",
       "I_Granite_11                   NaN         2.5147          4.7059   \n",
       "I_Obsidian_09               2.5147            NaN          3.2059   \n",
       "I_Pegmatite_01              4.7059         3.2059             NaN   \n",
       "I_Peridotite_11             3.6618         3.2059          5.7059   \n",
       "I_Pumice_06                 3.5000         2.3676          3.8382   \n",
       "I_Rhyolite_11               3.4706         3.9265          2.7059   \n",
       "M_Amphibolite_07            5.6029         4.7353          4.0441   \n",
       "M_Anthracite_12             3.0735         8.0000          3.3235   \n",
       "M_Gneiss_07                 3.8235         4.0735          4.6471   \n",
       "M_Hornfels_08               3.7206         5.5147          3.9559   \n",
       "M_Marble_04                 5.1765         3.3382          2.5588   \n",
       "M_Migmatite_08              4.8676         4.0000          4.8235   \n",
       "M_Phyllite_09               3.1912         4.3382          4.3382   \n",
       "M_Quartzite_09              4.9265         3.3235          3.1618   \n",
       "M_Schist_02                 5.0735         4.0294          5.4559   \n",
       "M_Slate_09                  2.8824         3.2500          2.8824   \n",
       "S_Bituminous Coal_10        3.2353         7.0588          3.2353   \n",
       "S_Breccia_11                4.3235         2.3235          2.8971   \n",
       "S_Chert_03                  3.0000         4.1912          4.2941   \n",
       "S_Conglomerate_08           3.9853         2.9559          5.8676   \n",
       "S_Dolomite_12               4.1618         3.2059          2.7941   \n",
       "S_Micrite_09                3.1618         4.4118          3.5000   \n",
       "S_Rock Gypsum_02            4.3676         4.0441          2.9118   \n",
       "S_Rock Salt_12              4.3235         3.7794          3.3971   \n",
       "S_Sandstone_12              2.5882         2.4706          2.7059   \n",
       "S_Shale_04                  2.7500         5.6029          2.7059   \n",
       "\n",
       "                      I_Peridotite_11  I_Pumice_06  I_Rhyolite_11  ...  \\\n",
       "I_Andesite_06                  5.1765       3.1912         5.5441  ...   \n",
       "I_Basalt_02                    3.9706       3.1765         6.5441  ...   \n",
       "I_Diorite_10                   5.3529       3.8529         3.5735  ...   \n",
       "I_Gabbro_03                    4.7206       3.3971         4.0735  ...   \n",
       "I_Granite_11                   3.6618       3.5000         3.4706  ...   \n",
       "I_Obsidian_09                  3.2059       2.3676         3.9265  ...   \n",
       "I_Pegmatite_01                 5.7059       3.8382         2.7059  ...   \n",
       "I_Peridotite_11                   NaN       5.5735         3.1765  ...   \n",
       "I_Pumice_06                    5.5735          NaN         3.8382  ...   \n",
       "I_Rhyolite_11                  3.1765       3.8382            NaN  ...   \n",
       "M_Amphibolite_07               4.8824       3.0882         4.2647  ...   \n",
       "M_Anthracite_12                2.9265       2.3971         3.7059  ...   \n",
       "M_Gneiss_07                    4.5147       4.5000         4.5588  ...   \n",
       "M_Hornfels_08                  4.4559       3.0882         5.8824  ...   \n",
       "M_Marble_04                    2.6471       2.9412         5.5000  ...   \n",
       "M_Migmatite_08                 5.5882       4.1176         4.3088  ...   \n",
       "M_Phyllite_09                  6.7206       4.5588         4.5000  ...   \n",
       "M_Quartzite_09                 3.4118       2.9706         5.8382  ...   \n",
       "M_Schist_02                    5.5441       3.6618         5.4265  ...   \n",
       "M_Slate_09                     2.8088       2.9412         6.5147  ...   \n",
       "S_Bituminous Coal_10           3.4265       2.9706         4.8382  ...   \n",
       "S_Breccia_11                   2.1765       2.1029         2.3235  ...   \n",
       "S_Chert_03                     3.3382       4.3088         3.8529  ...   \n",
       "S_Conglomerate_08              6.1471       4.5441         2.9559  ...   \n",
       "S_Dolomite_12                  3.3676       4.3382         6.9853  ...   \n",
       "S_Micrite_09                   4.1618       4.2500         4.4559  ...   \n",
       "S_Rock Gypsum_02               2.7647       2.9706         4.9853  ...   \n",
       "S_Rock Salt_12                 2.5294       2.1324         3.7794  ...   \n",
       "S_Sandstone_12                 2.9559       5.5147         3.9118  ...   \n",
       "S_Shale_04                     3.2941       2.3382         5.5294  ...   \n",
       "\n",
       "                      S_Bituminous Coal_10  S_Breccia_11  S_Chert_03  \\\n",
       "I_Andesite_06                       4.0294        4.2059      2.7941   \n",
       "I_Basalt_02                         5.7353        2.6765      3.1324   \n",
       "I_Diorite_10                        3.7206        3.2353      3.2353   \n",
       "I_Gabbro_03                         4.9559        4.0588      2.8382   \n",
       "I_Granite_11                        3.2353        4.3235      3.0000   \n",
       "I_Obsidian_09                       7.0588        2.3235      4.1912   \n",
       "I_Pegmatite_01                      3.2353        2.8971      4.2941   \n",
       "I_Peridotite_11                     3.4265        2.1765      3.3382   \n",
       "I_Pumice_06                         2.9706        2.1029      4.3088   \n",
       "I_Rhyolite_11                       4.8382        2.3235      3.8529   \n",
       "M_Amphibolite_07                    5.0588        3.8235      2.9853   \n",
       "M_Anthracite_12                     6.8824        2.4412      3.0147   \n",
       "M_Gneiss_07                         4.5735        2.7059      3.9706   \n",
       "M_Hornfels_08                       6.8382        3.1618      3.1176   \n",
       "M_Marble_04                         3.0882        3.7794      3.1618   \n",
       "M_Migmatite_08                      5.0294        2.2941      3.5735   \n",
       "M_Phyllite_09                       4.6765        2.1765      3.8529   \n",
       "M_Quartzite_09                      3.3529        4.1912      3.0294   \n",
       "M_Schist_02                         4.7059        3.1912      3.3235   \n",
       "M_Slate_09                          3.8676        2.0588      3.7941   \n",
       "S_Bituminous Coal_10                   NaN        2.2647      3.0882   \n",
       "S_Breccia_11                        2.2647           NaN      2.1912   \n",
       "S_Chert_03                          3.0882        2.1912         NaN   \n",
       "S_Conglomerate_08                   2.9706        2.8382      3.8088   \n",
       "S_Dolomite_12                       4.3971        2.5147      4.3235   \n",
       "S_Micrite_09                        4.7941        2.7353      5.7647   \n",
       "S_Rock Gypsum_02                    3.0147        3.2206      4.4853   \n",
       "S_Rock Salt_12                      3.2353        2.6765      2.9265   \n",
       "S_Sandstone_12                      3.0441        2.1471      4.6618   \n",
       "S_Shale_04                          6.8382        2.1618      3.2794   \n",
       "\n",
       "                      S_Conglomerate_08  S_Dolomite_12  S_Micrite_09  \\\n",
       "I_Andesite_06                    3.4118         4.1912        2.8676   \n",
       "I_Basalt_02                      2.6029         5.7500        6.1618   \n",
       "I_Diorite_10                     4.5000         4.6176        3.7353   \n",
       "I_Gabbro_03                      3.3676         4.1324        3.6176   \n",
       "I_Granite_11                     3.9853         4.1618        3.1618   \n",
       "I_Obsidian_09                    2.9559         3.2059        4.4118   \n",
       "I_Pegmatite_01                   5.8676         2.7941        3.5000   \n",
       "I_Peridotite_11                  6.1471         3.3676        4.1618   \n",
       "I_Pumice_06                      4.5441         4.3382        4.2500   \n",
       "I_Rhyolite_11                    2.9559         6.9853        4.4559   \n",
       "M_Amphibolite_07                 3.4412         3.1765        2.9265   \n",
       "M_Anthracite_12                  2.6618         2.9118        3.9118   \n",
       "M_Gneiss_07                      4.3382         4.4559        5.3824   \n",
       "M_Hornfels_08                    3.6765         4.7941        5.5000   \n",
       "M_Marble_04                      2.4559         5.5441        3.9412   \n",
       "M_Migmatite_08                   4.2206         5.1618        4.4853   \n",
       "M_Phyllite_09                    4.5147         4.3676        5.4559   \n",
       "M_Quartzite_09                   2.9412         5.1912        3.8382   \n",
       "M_Schist_02                      5.2206         4.8529        3.7206   \n",
       "M_Slate_09                       2.7941         5.4706        4.2059   \n",
       "S_Bituminous Coal_10             2.9706         4.3971        4.7941   \n",
       "S_Breccia_11                     2.8382         2.5147        2.7353   \n",
       "S_Chert_03                       3.8088         4.3235        5.7647   \n",
       "S_Conglomerate_08                   NaN         2.8088        4.4265   \n",
       "S_Dolomite_12                    2.8088            NaN        5.7059   \n",
       "S_Micrite_09                     4.4265         5.7059           NaN   \n",
       "S_Rock Gypsum_02                 2.1912         5.3235        4.3235   \n",
       "S_Rock Salt_12                   2.6765         3.7794        2.5882   \n",
       "S_Sandstone_12                   3.1029         3.9412        4.0588   \n",
       "S_Shale_04                       2.6471         5.5882        5.1765   \n",
       "\n",
       "                      S_Rock Gypsum_02  S_Rock Salt_12  S_Sandstone_12  \\\n",
       "I_Andesite_06                   3.3676          5.0000          2.4118   \n",
       "I_Basalt_02                     4.6765          3.8824          3.2941   \n",
       "I_Diorite_10                    3.2647          3.7500          2.6471   \n",
       "I_Gabbro_03                     3.7794          3.2206          2.2794   \n",
       "I_Granite_11                    4.3676          4.3235          2.5882   \n",
       "I_Obsidian_09                   4.0441          3.7794          2.4706   \n",
       "I_Pegmatite_01                  2.9118          3.3971          2.7059   \n",
       "I_Peridotite_11                 2.7647          2.5294          2.9559   \n",
       "I_Pumice_06                     2.9706          2.1324          5.5147   \n",
       "I_Rhyolite_11                   4.9853          3.7794          3.9118   \n",
       "M_Amphibolite_07                3.3971          4.6765          2.3382   \n",
       "M_Anthracite_12                 3.5588          4.5000          2.1324   \n",
       "M_Gneiss_07                     3.2794          2.5735          3.2794   \n",
       "M_Hornfels_08                   3.9706          3.6618          2.8088   \n",
       "M_Marble_04                     7.0882          6.4706          2.6176   \n",
       "M_Migmatite_08                  3.0735          2.9265          3.5588   \n",
       "M_Phyllite_09                   3.0882          2.9265          3.9706   \n",
       "M_Quartzite_09                  5.5882          5.0147          2.6176   \n",
       "M_Schist_02                     3.8529          5.0882          2.6324   \n",
       "M_Slate_09                      4.1029          3.5588          2.7059   \n",
       "S_Bituminous Coal_10            3.0147          3.2353          3.0441   \n",
       "S_Breccia_11                    3.2206          2.6765          2.1471   \n",
       "S_Chert_03                      4.4853          2.9265          4.6618   \n",
       "S_Conglomerate_08               2.1912          2.6765          3.1029   \n",
       "S_Dolomite_12                   5.3235          3.7794          3.9412   \n",
       "S_Micrite_09                    4.3235          2.5882          4.0588   \n",
       "S_Rock Gypsum_02                   NaN          5.7794          2.5294   \n",
       "S_Rock Salt_12                  5.7794             NaN          1.8676   \n",
       "S_Sandstone_12                  2.5294          1.8676             NaN   \n",
       "S_Shale_04                      3.1029          2.9412          2.7794   \n",
       "\n",
       "                      S_Shale_04  \n",
       "I_Andesite_06             4.7206  \n",
       "I_Basalt_02               6.7353  \n",
       "I_Diorite_10              3.7059  \n",
       "I_Gabbro_03               5.0294  \n",
       "I_Granite_11              2.7500  \n",
       "I_Obsidian_09             5.6029  \n",
       "I_Pegmatite_01            2.7059  \n",
       "I_Peridotite_11           3.2941  \n",
       "I_Pumice_06               2.3382  \n",
       "I_Rhyolite_11             5.5294  \n",
       "M_Amphibolite_07          4.9412  \n",
       "M_Anthracite_12           5.2500  \n",
       "M_Gneiss_07               4.6324  \n",
       "M_Hornfels_08             5.3971  \n",
       "M_Marble_04               2.9118  \n",
       "M_Migmatite_08            4.4559  \n",
       "M_Phyllite_09             3.8088  \n",
       "M_Quartzite_09            4.0294  \n",
       "M_Schist_02               3.9412  \n",
       "M_Slate_09                5.8088  \n",
       "S_Bituminous Coal_10      6.8382  \n",
       "S_Breccia_11              2.1618  \n",
       "S_Chert_03                3.2794  \n",
       "S_Conglomerate_08         2.6471  \n",
       "S_Dolomite_12             5.5882  \n",
       "S_Micrite_09              5.1765  \n",
       "S_Rock Gypsum_02          3.1029  \n",
       "S_Rock Salt_12            2.9412  \n",
       "S_Sandstone_12            2.7794  \n",
       "S_Shale_04                   NaN  \n",
       "\n",
       "[30 rows x 30 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "human_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "client = openai.OpenAI(\n",
    "  organization='org-LewILTfUI8Y9pb57686FCqdy',\n",
    "  project='proj_fPgG23M7oU6YDmEKQRAc5H7y',\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Function to encode the image\n",
    "def encode_image(image):\n",
    "  image_path = f\"data\\\\360 Rocks\\\\{image}.jpg\"\n",
    "  with open(image_path, \"rb\") as image_file:\n",
    "    return base64.b64encode(image_file.read()).decode('utf-8')  \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "SYSTEM_PROMPT = (\n",
    "    \"You are assisting in a study in which you are shown pairs of rocks and \"\n",
    "    \"rate how visually similar they are on a scale from 1 to 9, with 1 being most dissimilar, \"\n",
    "    \"5 being moderately similar, and 9 being most similar. \"\n",
    "    \"You are shown examples of highly similar, moderately similar, and highly dissimilar pairs. \"\n",
    "    \"You use these examples to anchor your judgments. \"\n",
    "    # ChatGPT is biased to give low similarities. Adding this instruction helps.\n",
    "    \"You use the full range of the scale and only respond with a 1 or 2 when the pair is extremely dissimilar. \"\n",
    "    \"You only respond with a single number from 1 to 9, without explaining your reasoning.\"\n",
    "    )\n",
    "\n",
    "SIMILAR_EXAMPLE_PROMPT = (\n",
    "    \"This is an example of a highly similar pair of rocks.\"\n",
    ")\n",
    "\n",
    "DISSIMILAR_EXAMPLE_PROMPT = (\n",
    "    \"This is an example of a highly dissimilar pair of rocks.\"\n",
    ")\n",
    "\n",
    "MEDIUM_EXAMPLE_PROMPT = (\n",
    "    \"This is an example of a moderately similar pair of rocks.\"\n",
    ")\n",
    "\n",
    "QUESTION_PROMPT = (\n",
    "    \"From 1-9, how visually similar are these two rocks?\"\n",
    ")\n",
    "\n",
    "SIMILAR_ROCK1 = \"I_Obsidian_09\"\n",
    "SIMILAR_ROCK2 = \"M_Anthracite_12\"\n",
    "\n",
    "DISSIMILAR_ROCK1 = \"S_Rock Salt_12\"\n",
    "DISSIMILAR_ROCK2 = \"S_Sandstone_12\"\n",
    "\n",
    "MEDIUM_ROCK1 = \"I_Andesite_06\"\n",
    "MEDIUM_ROCK2 = \"I_Peridotite_11\"\n",
    "\n",
    "SEED = 123"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_messages(prompt, rocks):\n",
    "    messages = {\"role\": \"user\",\n",
    "       \"content\": [{\"type\": \"text\", \"text\": prompt}] + [\n",
    "         {\n",
    "            \"type\": \"image_url\",\n",
    "            \"image_url\": {\n",
    "              \"url\": f\"data:image/jpeg;base64,{encode_image(rock)}\"\n",
    "            },\n",
    "          }\n",
    "        for rock in rocks]\n",
    "    }\n",
    "    return messages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "@retry(wait=wait_random_exponential(min=1, max=60), stop=stop_after_attempt(10))\n",
    "def completion_with_backoff(**kwargs):\n",
    "    return client.chat.completions.create(**kwargs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_responses(rock1, rock2): \n",
    "  response = completion_with_backoff(\n",
    "    model=\"gpt-4o\",\n",
    "    messages=[\n",
    "      {\"role\": \"system\", \"content\": SYSTEM_PROMPT},\n",
    "      create_messages(SIMILAR_EXAMPLE_PROMPT, (SIMILAR_ROCK1, SIMILAR_ROCK2)),\n",
    "      create_messages(MEDIUM_EXAMPLE_PROMPT, (MEDIUM_ROCK1, MEDIUM_ROCK2)),\n",
    "      create_messages(DISSIMILAR_EXAMPLE_PROMPT, (DISSIMILAR_ROCK1, DISSIMILAR_ROCK2)),\n",
    "      create_messages(QUESTION_PROMPT, (rock1, rock2)),\n",
    "    ],\n",
    "    logprobs=True,\n",
    "    top_logprobs=20,\n",
    "    seed=SEED,\n",
    "    max_tokens=1,\n",
    "  )\n",
    "  return response"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_average_rating(logprobs):\n",
    "    # We get a weighted average using the logprobs. The logprobs are actually nondeterministic\n",
    "    # even with a fixed random seed, so different runs might have slightly different results.\n",
    "    word2num = {'one': 1, 'two': 2, 'three': 3, 'four': 4, 'five': 5, 'six': 6, 'seven': 7, 'eight': 8, 'nine': 9}\n",
    "    ratings = np.array([i+1 for i in range(9)])\n",
    "    weights = np.array([0.]*9)\n",
    "    for lp in logprobs:\n",
    "        token = lp.token.strip().lower()\n",
    "        i = word2num.get(token, token) # Convert what we can to numerical digits\n",
    "        try:\n",
    "            i = int(i) - 1\n",
    "            weights[i] += np.exp(lp.logprob)\n",
    "        except:\n",
    "            continue\n",
    "    try:\n",
    "        av = np.average(ratings, weights=weights)\n",
    "    except:\n",
    "        av = 0.0\n",
    "    return av\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "results = []\n",
    "pairs = [comb for comb in combinations(human_data.columns, 2)]\n",
    "for pair in pairs:\n",
    "    rock1, rock2 = pair\n",
    "\n",
    "    human = human_data[rock1][rock2]\n",
    "\n",
    "    response = get_responses(rock1, rock2)\n",
    "    logprobs = response.choices[0].logprobs.content[0].top_logprobs\n",
    "    chatgpt = get_average_rating(logprobs)\n",
    "\n",
    "    results.append({\"Rock1\": rock1, \"Rock2\": rock2, \"Human Rating\": human, \"ChatGPT Rating\": chatgpt, \"Response\": response})\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "results = pd.DataFrame(results)\n",
    "results.to_csv(\"chatgpt_similarities.csv\", index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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>Rock1</th>\n",
       "      <th>Rock2</th>\n",
       "      <th>Human Rating</th>\n",
       "      <th>ChatGPT Rating</th>\n",
       "      <th>Response</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>I_Andesite_06</td>\n",
       "      <td>I_Basalt_02</td>\n",
       "      <td>5.5294</td>\n",
       "      <td>5.410093</td>\n",
       "      <td>ChatCompletion(id='chatcmpl-9UL8zfwHKVjE5ebF9O...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>I_Andesite_06</td>\n",
       "      <td>I_Diorite_10</td>\n",
       "      <td>6.7059</td>\n",
       "      <td>7.463376</td>\n",
       "      <td>ChatCompletion(id='chatcmpl-9UL95Bg5bc8Mp5hdaZ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>I_Andesite_06</td>\n",
       "      <td>I_Gabbro_03</td>\n",
       "      <td>6.8824</td>\n",
       "      <td>7.926699</td>\n",
       "      <td>ChatCompletion(id='chatcmpl-9UL9IbRfPQ6JtOsl0q...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>I_Andesite_06</td>\n",
       "      <td>I_Granite_11</td>\n",
       "      <td>5.6324</td>\n",
       "      <td>5.896600</td>\n",
       "      <td>ChatCompletion(id='chatcmpl-9UL9QoKCypO0ACu915...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>I_Andesite_06</td>\n",
       "      <td>I_Obsidian_09</td>\n",
       "      <td>3.7206</td>\n",
       "      <td>2.048678</td>\n",
       "      <td>ChatCompletion(id='chatcmpl-9UL9WQLaMTrNUGxtDc...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>430</th>\n",
       "      <td>S_Rock Gypsum_02</td>\n",
       "      <td>S_Sandstone_12</td>\n",
       "      <td>2.5294</td>\n",
       "      <td>1.895034</td>\n",
       "      <td>ChatCompletion(id='chatcmpl-9ULyNrj3hTItwCxGzf...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>431</th>\n",
       "      <td>S_Rock Gypsum_02</td>\n",
       "      <td>S_Shale_04</td>\n",
       "      <td>3.1029</td>\n",
       "      <td>1.968626</td>\n",
       "      <td>ChatCompletion(id='chatcmpl-9ULyW0NvmejOeF2nx3...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>432</th>\n",
       "      <td>S_Rock Salt_12</td>\n",
       "      <td>S_Sandstone_12</td>\n",
       "      <td>1.8676</td>\n",
       "      <td>1.033163</td>\n",
       "      <td>ChatCompletion(id='chatcmpl-9ULygUerSEhZQ3p6Yt...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>433</th>\n",
       "      <td>S_Rock Salt_12</td>\n",
       "      <td>S_Shale_04</td>\n",
       "      <td>2.9412</td>\n",
       "      <td>1.995280</td>\n",
       "      <td>ChatCompletion(id='chatcmpl-9ULymDucFo9kWMuFOV...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>434</th>\n",
       "      <td>S_Sandstone_12</td>\n",
       "      <td>S_Shale_04</td>\n",
       "      <td>2.7794</td>\n",
       "      <td>1.789263</td>\n",
       "      <td>ChatCompletion(id='chatcmpl-9ULys8KVx5FNqU3sl2...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>435 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                Rock1           Rock2  Human Rating  ChatGPT Rating  \\\n",
       "0       I_Andesite_06     I_Basalt_02        5.5294        5.410093   \n",
       "1       I_Andesite_06    I_Diorite_10        6.7059        7.463376   \n",
       "2       I_Andesite_06     I_Gabbro_03        6.8824        7.926699   \n",
       "3       I_Andesite_06    I_Granite_11        5.6324        5.896600   \n",
       "4       I_Andesite_06   I_Obsidian_09        3.7206        2.048678   \n",
       "..                ...             ...           ...             ...   \n",
       "430  S_Rock Gypsum_02  S_Sandstone_12        2.5294        1.895034   \n",
       "431  S_Rock Gypsum_02      S_Shale_04        3.1029        1.968626   \n",
       "432    S_Rock Salt_12  S_Sandstone_12        1.8676        1.033163   \n",
       "433    S_Rock Salt_12      S_Shale_04        2.9412        1.995280   \n",
       "434    S_Sandstone_12      S_Shale_04        2.7794        1.789263   \n",
       "\n",
       "                                              Response  \n",
       "0    ChatCompletion(id='chatcmpl-9UL8zfwHKVjE5ebF9O...  \n",
       "1    ChatCompletion(id='chatcmpl-9UL95Bg5bc8Mp5hdaZ...  \n",
       "2    ChatCompletion(id='chatcmpl-9UL9IbRfPQ6JtOsl0q...  \n",
       "3    ChatCompletion(id='chatcmpl-9UL9QoKCypO0ACu915...  \n",
       "4    ChatCompletion(id='chatcmpl-9UL9WQLaMTrNUGxtDc...  \n",
       "..                                                 ...  \n",
       "430  ChatCompletion(id='chatcmpl-9ULyNrj3hTItwCxGzf...  \n",
       "431  ChatCompletion(id='chatcmpl-9ULyW0NvmejOeF2nx3...  \n",
       "432  ChatCompletion(id='chatcmpl-9ULygUerSEhZQ3p6Yt...  \n",
       "433  ChatCompletion(id='chatcmpl-9ULymDucFo9kWMuFOV...  \n",
       "434  ChatCompletion(id='chatcmpl-9ULys8KVx5FNqU3sl2...  \n",
       "\n",
       "[435 rows x 5 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x1e34abae8f0>]"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.scatterplot(results, x=\"Human Rating\", y=\"ChatGPT Rating\")\n",
    "plt.plot([1, 9], [1, 9], 'k--')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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>Human Rating</th>\n",
       "      <th>ChatGPT Rating</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Human Rating</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.781791</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ChatGPT Rating</th>\n",
       "      <td>0.781791</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                Human Rating  ChatGPT Rating\n",
       "Human Rating        1.000000        0.781791\n",
       "ChatGPT Rating      0.781791        1.000000"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results.corr(numeric_only=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "results[\"Human Rating\"].hist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: >"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "results[\"ChatGPT Rating\"].hist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "base",
   "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": "3.10.11"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}