Add notebook performing deduplication of the master files

notebooks/dedupe_full_jiggins_DL.ipynb

@@ -0,0 +1,1955 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas as pd"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Summary\n",
+    "\n",
+    "We explored our downloaded images in `EDA_full_jiggins_DL.ipynb`, and established the following causes of duplication along with remedies:\n",
+    "1. The \"patch\" record (record 3477891) which required alignment in `record_3477891_media_occurrence.ipynb`, resulting in `Jiggins_Zenodo_Img_Master_3477891Patch.csv` used for download, is a duplication of all listed records. Additionally, it introduced extra copies of 4 images with incorrect labels for their views (i.e., dorsal vs ventral).\n",
+    " - Solution: This record will be removed.\n",
+    "2. The remaining duplication (430 images x2) comes from 5 records: [4291095](https://zenodo.org/records/4291095), [2813153](https://zenodo.org/records/2813153), [5526257](https://zenodo.org/records/5526257), [2553977](https://zenodo.org/records/2553977), and [2552371](https://zenodo.org/records/2552371).\n",
+    "  - Though there is nothing in their descriptions on Zenodo to indicate this overlap occurred, record 4291095 duplicated 415 images from record 2813153. 104 of these images had their view/side labels (dorsal vs ventral) updated to reflect an earlier mislabeling in record 2813153; these were all RAW copies of the images. All other metadata was consistent across records.\n",
+    "  - Solution: Copies of these images and their metadata will be retained from the newer record, record 4291095.\n",
+    "  - Record 5526257 has 10 images that were added twice. All their metadata is consistent, so we will just keep the first instance of each image.\n",
+    "  - Records 2553977 and 2552371 are both \"Miscellaneous Heliconius wing photographs (2001-2019)\" Parts 3 & 1, respectively. Record 2553977 has duplicated images with matching metadata, though 3 of the 4 images have different filenames (`Image_name`). It seems to have resulted from typos or not indicating `cut`, as they are all close-up crops of a single wing. From this perspective, the view labels (dorsal) don't seem appropriate when we have more fine-grained indicators such as \"forewing dorsal\".\n",
+    "    - Suggestion that the entries in this record with `cut` in the filename not be used for general classification unless this variety is desired. These also appear to be all the `tif` images, which are often ignored.\n",
+    "    - Record 2552371 has one image duplicated with two different `CAMID`s assigned (different views as well). These will both be removed as the image has no indicator of the specimen ID in it. Note that these also were labeled `Heliconius sp.`.\n",
+    "3. Looking at images that were not duplicates, there are clearly still multiple images of the same specimen from the same perspective (eg., two dorsal images) that are also of the same file type (eg., both jpgs). Duplication was checked at the pixel level, so there is no guarantee that they are truly different images, but due to the scope of this collection (over 20 years), it does not seem unlikely that multiple images could have been taken of the same specimen. \n",
+    "  - Suggestion/Solution: When determining splits using this data, follow these steps:\n",
+    "    1. Ensure you are looking at only **_one_** view (eg., dorsal or ventral).\n",
+    "    2. Reduce to only unique `CAMID`s.\n",
+    "    3. Generate splits as desired.\n",
+    "    4. Add images to splits based on matched `CAMID`s if multiple views are desired (i.e., if a `CAMID` is present in the dorsal training, add the matching ventral image to the training set).\n",
+    "\n",
+    "\n",
+    "## Generate \"Clean\" Master Metadata File\n",
+    "\n",
+    "We'll now generate a cleaned master metadata file from `metadata/Jiggins_Zenodo_Img_Master_3477891Patch_downloaded.csv` (generated in `notebooks/EDA_full_jiggins_DL.ipynb` from our download master file and checksums of successful downloads) following the solutions outlined above. "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "STATS_COLS = [\"X\",\n",
+    "              \"CAMID\",\n",
+    "              \"Image_name\",\n",
+    "              \"md5\",\n",
+    "              \"record_number\",\n",
+    "              \"Taxonomic_Name\",\n",
+    "              \"species\",\n",
+    "              \"subspecies\",\n",
+    "              \"genus\",\n",
+    "              \"Cross_Type\",\n",
+    "              \"View\",\n",
+    "              \"Locality\",\n",
+    "              \"Sex\",\n",
+    "              \"Dataset\",\n",
+    "              \"file_type\"]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "<class 'pandas.core.frame.DataFrame'>\n",
+      "RangeIndex: 42143 entries, 0 to 42142\n",
+      "Data columns (total 15 columns):\n",
+      " #   Column          Non-Null Count  Dtype \n",
+      "---  ------          --------------  ----- \n",
+      " 0   X               42143 non-null  int64 \n",
+      " 1   CAMID           42143 non-null  object\n",
+      " 2   Image_name      42143 non-null  object\n",
+      " 3   md5             42143 non-null  object\n",
+      " 4   record_number   42143 non-null  int64 \n",
+      " 5   Taxonomic_Name  42143 non-null  object\n",
+      " 6   species         42143 non-null  object\n",
+      " 7   subspecies      24502 non-null  object\n",
+      " 8   genus           42143 non-null  object\n",
+      " 9   Cross_Type      4447 non-null   object\n",
+      " 10  View            41364 non-null  object\n",
+      " 11  Locality        29047 non-null  object\n",
+      " 12  Sex             33254 non-null  object\n",
+      " 13  Dataset         34924 non-null  object\n",
+      " 14  file_type       42143 non-null  object\n",
+      "dtypes: int64(2), object(13)\n",
+      "memory usage: 4.8+ MB\n"
+     ]
+    }
+   ],
+   "source": [
+    "df = pd.read_csv(\"../metadata/Jiggins_Zenodo_Img_Master_3477891Patch_downloaded.csv\", low_memory = False)\n",
+    "df[STATS_COLS].info()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "X                 42143\n",
+       "CAMID             11968\n",
+       "Image_name        36216\n",
+       "md5               36212\n",
+       "record_number        30\n",
+       "Taxonomic_Name      366\n",
+       "species             246\n",
+       "subspecies          155\n",
+       "genus                94\n",
+       "Cross_Type           30\n",
+       "View                  7\n",
+       "Locality            644\n",
+       "Sex                   3\n",
+       "Dataset               7\n",
+       "file_type             3\n",
+       "dtype: int64"
+      ]
+     },
+     "execution_count": 4,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "df[STATS_COLS].nunique()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Observations:\n",
+    "- Half of the Zenodo records wind up with duplicated images. \n",
+    "- `Image_name` uniqueness is close to `md5`, but there are 4 more unique image names than unique images. I believe this was from the duplication in reocrd 2553977. We can check when cleaning it. \n",
+    "- None of the cross types are duplicated.\n",
+    "- We do have multiple unique images for a single `CAMID`, more than just dorsal vs ventral since there are 11968 unique CAMIDs and 36212 unique `md5` values (36212 = **3**x11968 + 308)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Remove Record 3477891\n",
+    "\n",
+    "This is just a duplication of images and we already have from earlier records, along with some extra duplication that gets mislabeled."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "X                 36642\n",
+       "CAMID             11968\n",
+       "Image_name        36216\n",
+       "md5               36212\n",
+       "record_number        29\n",
+       "Taxonomic_Name      366\n",
+       "species             246\n",
+       "subspecies          155\n",
+       "genus                94\n",
+       "Cross_Type           30\n",
+       "View                  7\n",
+       "Locality            644\n",
+       "Sex                   3\n",
+       "Dataset               7\n",
+       "file_type             3\n",
+       "dtype: int64"
+      ]
+     },
+     "execution_count": 5,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "df_reduced = df.loc[df[\"record_number\"] != 3477891].copy()\n",
+    "df_reduced[STATS_COLS].nunique()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "The only change is in `X` (our total entry count) and in `record_number` (expected since we removed one)."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Address Remaining Duplicates\n",
+    "\n",
+    "The remaining duplicates are across 5 records and will require a bit more care to reslove (following the plan outlined at the top of the notebook and in the README).\n",
+    "\n",
+    "### Record 4291095 & 2813153 Duplicates\n",
+    "\n",
+    "There are 415 duplicates across these two records, with `View` corrections in record 4291095, so we will remove the duplicates from record 2813153."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "Index(['CAMID', 'X', 'Image_name', 'View', 'zenodo_name', 'zenodo_link',\n",
+       "       'Sequence', 'Taxonomic_Name', 'Locality', 'Sample_accession',\n",
+       "       'Collected_by', 'Other_ID', 'Date', 'Dataset', 'Store', 'Brood',\n",
+       "       'Death_Date', 'Cross_Type', 'Stage', 'Sex', 'Unit_Type', 'file_type',\n",
+       "       'record_number', 'species', 'subspecies', 'genus', 'file_url',\n",
+       "       'hybrid_stat', 'filename', 'filepath', 'md5'],\n",
+       "      dtype='object')"
+      ]
+     },
+     "execution_count": 6,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "df_reduced.columns"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "We start by marking all duplicate images since the decision of which to keep is dependent on comparisons between them."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "md5_duplicated\n",
+       "False    35782\n",
+       "True       860\n",
+       "Name: count, dtype: int64"
+      ]
+     },
+     "execution_count": 7,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "df_reduced[\"md5_duplicated\"] = df_reduced.duplicated(\"md5\", keep = False)\n",
+    "df_reduced[\"md5_duplicated\"].value_counts()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "415"
+      ]
+     },
+     "execution_count": 8,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "duplicates = df_reduced[df_reduced[\"md5_duplicated\"]]\n",
+    "dupes_2813153 = duplicates.loc[duplicates[\"record_number\"] == 2813153]\n",
+    "#print(duplicates.loc[duplicates[\"record_number\"] == 4291095])\n",
+    "dupes_2813153.shape[0]"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "We will record the `X` value from these 415 entries to remove them from the overall dataset in favor of the updated entries from record 4291095."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 9,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "X_dupes_2813153 = list(dupes_2813153.X)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "X                 36227\n",
+       "CAMID             11968\n",
+       "Image_name        36216\n",
+       "md5               36212\n",
+       "record_number        29\n",
+       "Taxonomic_Name      366\n",
+       "species             246\n",
+       "subspecies          155\n",
+       "genus                94\n",
+       "Cross_Type           30\n",
+       "View                  7\n",
+       "Locality            644\n",
+       "Sex                   3\n",
+       "Dataset               7\n",
+       "file_type             3\n",
+       "dtype: int64"
+      ]
+     },
+     "execution_count": 10,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "df_reduced_2 = df_reduced.loc[~df_reduced[\"X\"].isin(X_dupes_2813153)]\n",
+    "df_reduced_2[STATS_COLS].nunique()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "As expected, our numbers are holding steady with a reduction in `X` values of 415."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Record 5526257 Duplicates\n",
+    "\n",
+    "For record 5526257, the 10 duplicated entries are all the same, so we will just check `.duplicated` and keep the first instance of each."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "dupe_entries\n",
+      "False    10\n",
+      "True     10\n",
+      "Name: count, dtype: int64\n"
+     ]
+    },
+    {
+     "data": {
+      "text/plain": [
+       "10"
+      ]
+     },
+     "execution_count": 11,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "dupes_5526257 = duplicates.loc[duplicates[\"record_number\"] == 5526257].copy()\n",
+    "dupes_5526257[\"dupe_entries\"] = dupes_5526257.duplicated([\"CAMID\", \"Image_name\", \"View\", \"Taxonomic_Name\", \"Locality\", \"md5\"], keep = \"first\")\n",
+    "print(dupes_5526257[\"dupe_entries\"].value_counts())\n",
+    "\n",
+    "X_dupes_5526257 = list(dupes_5526257.loc[dupes_5526257[\"dupe_entries\"], \"X\"])\n",
+    "len(X_dupes_5526257)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 12,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "X                 36217\n",
+       "CAMID             11968\n",
+       "Image_name        36216\n",
+       "md5               36212\n",
+       "record_number        29\n",
+       "Taxonomic_Name      366\n",
+       "species             246\n",
+       "subspecies          155\n",
+       "genus                94\n",
+       "Cross_Type           30\n",
+       "View                  7\n",
+       "Locality            644\n",
+       "Sex                   3\n",
+       "Dataset               7\n",
+       "file_type             3\n",
+       "dtype: int64"
+      ]
+     },
+     "execution_count": 12,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "df_reduced_3 = df_reduced_2.loc[~df_reduced_2[\"X\"].isin(X_dupes_5526257)]\n",
+    "df_reduced_3[STATS_COLS].nunique()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Reduction in `X` by 10, consistency in all other numbers.\n",
+    "\n",
+    "### Record 2552371 Duplicates\n",
+    "\n",
+    "We are removing both duplicates from this record as their metadata is contradictory and cannot be verified. (Two different CAMIDs, different filenames, different views, etc.)"
+   ]
+  },
+  {
+   "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>CAMID</th>\n",
+       "      <th>X</th>\n",
+       "      <th>Image_name</th>\n",
+       "      <th>View</th>\n",
+       "      <th>zenodo_name</th>\n",
+       "      <th>zenodo_link</th>\n",
+       "      <th>Sequence</th>\n",
+       "      <th>Taxonomic_Name</th>\n",
+       "      <th>Locality</th>\n",
+       "      <th>Sample_accession</th>\n",
+       "      <th>...</th>\n",
+       "      <th>record_number</th>\n",
+       "      <th>species</th>\n",
+       "      <th>subspecies</th>\n",
+       "      <th>genus</th>\n",
+       "      <th>file_url</th>\n",
+       "      <th>hybrid_stat</th>\n",
+       "      <th>filename</th>\n",
+       "      <th>filepath</th>\n",
+       "      <th>md5</th>\n",
+       "      <th>md5_duplicated</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>11978</th>\n",
+       "      <td>CAM010354</td>\n",
+       "      <td>23471</td>\n",
+       "      <td>10354d.jpg</td>\n",
+       "      <td>dorsal</td>\n",
+       "      <td>Heliconius_wing_old_photos_2001_2019_part1.csv</td>\n",
+       "      <td>https://zenodo.org/record/2552371</td>\n",
+       "      <td>10,354</td>\n",
+       "      <td>Heliconius sp.</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>...</td>\n",
+       "      <td>2552371</td>\n",
+       "      <td>Heliconius sp.</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>Heliconius</td>\n",
+       "      <td>https://zenodo.org/record/2552371/files/10354d...</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>23471_10354d.jpg</td>\n",
+       "      <td>images/Heliconius sp./23471_10354d.jpg</td>\n",
+       "      <td>84d4ac6527458786cdb33166cd80e0a8</td>\n",
+       "      <td>True</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>12010</th>\n",
+       "      <td>CAM010362</td>\n",
+       "      <td>23484</td>\n",
+       "      <td>10362v.jpg</td>\n",
+       "      <td>ventral</td>\n",
+       "      <td>Heliconius_wing_old_photos_2001_2019_part1.csv</td>\n",
+       "      <td>https://zenodo.org/record/2552371</td>\n",
+       "      <td>10,362</td>\n",
+       "      <td>Heliconius sp.</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>...</td>\n",
+       "      <td>2552371</td>\n",
+       "      <td>Heliconius sp.</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>Heliconius</td>\n",
+       "      <td>https://zenodo.org/record/2552371/files/10362v...</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>23484_10362v.jpg</td>\n",
+       "      <td>images/Heliconius sp./23484_10362v.jpg</td>\n",
+       "      <td>84d4ac6527458786cdb33166cd80e0a8</td>\n",
+       "      <td>True</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>2 rows × 32 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "           CAMID      X  Image_name     View  \\\n",
+       "11978  CAM010354  23471  10354d.jpg   dorsal   \n",
+       "12010  CAM010362  23484  10362v.jpg  ventral   \n",
+       "\n",
+       "                                          zenodo_name  \\\n",
+       "11978  Heliconius_wing_old_photos_2001_2019_part1.csv   \n",
+       "12010  Heliconius_wing_old_photos_2001_2019_part1.csv   \n",
+       "\n",
+       "                             zenodo_link Sequence  Taxonomic_Name Locality  \\\n",
+       "11978  https://zenodo.org/record/2552371   10,354  Heliconius sp.      NaN   \n",
+       "12010  https://zenodo.org/record/2552371   10,362  Heliconius sp.      NaN   \n",
+       "\n",
+       "      Sample_accession  ... record_number         species subspecies  \\\n",
+       "11978              NaN  ...       2552371  Heliconius sp.        NaN   \n",
+       "12010              NaN  ...       2552371  Heliconius sp.        NaN   \n",
+       "\n",
+       "            genus                                           file_url  \\\n",
+       "11978  Heliconius  https://zenodo.org/record/2552371/files/10354d...   \n",
+       "12010  Heliconius  https://zenodo.org/record/2552371/files/10362v...   \n",
+       "\n",
+       "      hybrid_stat          filename                                filepath  \\\n",
+       "11978         NaN  23471_10354d.jpg  images/Heliconius sp./23471_10354d.jpg   \n",
+       "12010         NaN  23484_10362v.jpg  images/Heliconius sp./23484_10362v.jpg   \n",
+       "\n",
+       "                                    md5 md5_duplicated  \n",
+       "11978  84d4ac6527458786cdb33166cd80e0a8           True  \n",
+       "12010  84d4ac6527458786cdb33166cd80e0a8           True  \n",
+       "\n",
+       "[2 rows x 32 columns]"
+      ]
+     },
+     "execution_count": 13,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "dupes_2552371 = duplicates.loc[duplicates[\"record_number\"] == 2552371]\n",
+    "X_dupes_2552371 = list(dupes_2552371[\"X\"])\n",
+    "dupes_2552371"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 14,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "X                 36215\n",
+       "CAMID             11968\n",
+       "Image_name        36214\n",
+       "md5               36211\n",
+       "record_number        29\n",
+       "Taxonomic_Name      366\n",
+       "species             246\n",
+       "subspecies          155\n",
+       "genus                94\n",
+       "Cross_Type           30\n",
+       "View                  7\n",
+       "Locality            644\n",
+       "Sex                   3\n",
+       "Dataset               7\n",
+       "file_type             3\n",
+       "dtype: int64"
+      ]
+     },
+     "execution_count": 14,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "df_reduced_4 = df_reduced_3.loc[~df_reduced_3[\"X\"].isin(X_dupes_2552371)]\n",
+    "df_reduced_4[STATS_COLS].nunique()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Removing these two has reduced our `X` count by 2, as well as our number of unique `Image_name`s, and it reduced our unique image count (`md5`) by 1. Observe that our unique `CAMID` count has remained constant, so both specimens indicated by these duplicate images are still represented in the dataset."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Record 2553977 Duplicates\n",
+    "\n",
+    "The duplicated entries for record 2553977 are the most complicated in that their metadata is nearly exact, though a few have different filenames (`Image_name`) and we would like to preserve the version that has `_cut` in the name."
+   ]
+  },
+  {
+   "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>X</th>\n",
+       "      <th>CAMID</th>\n",
+       "      <th>Image_name</th>\n",
+       "      <th>md5</th>\n",
+       "      <th>record_number</th>\n",
+       "      <th>Taxonomic_Name</th>\n",
+       "      <th>species</th>\n",
+       "      <th>subspecies</th>\n",
+       "      <th>genus</th>\n",
+       "      <th>Cross_Type</th>\n",
+       "      <th>View</th>\n",
+       "      <th>Locality</th>\n",
+       "      <th>Sex</th>\n",
+       "      <th>Dataset</th>\n",
+       "      <th>file_type</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>40141</th>\n",
+       "      <td>26197</td>\n",
+       "      <td>CAM050147</td>\n",
+       "      <td>CAM050147_DS1_HW_IMG_8537_cut_3.tif</td>\n",
+       "      <td>21e726df7076f5c13fe1441126bac4d5</td>\n",
+       "      <td>2553977</td>\n",
+       "      <td>Heliconius sara ssp. sara</td>\n",
+       "      <td>Heliconius sara</td>\n",
+       "      <td>sara</td>\n",
+       "      <td>Heliconius</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>dorsal</td>\n",
+       "      <td>Madingley</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>tif</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>40144</th>\n",
+       "      <td>26196</td>\n",
+       "      <td>CAM050147</td>\n",
+       "      <td>CAM050147_DS1_HW_IMG_8537_cut_3.tif</td>\n",
+       "      <td>21e726df7076f5c13fe1441126bac4d5</td>\n",
+       "      <td>2553977</td>\n",
+       "      <td>Heliconius sara ssp. sara</td>\n",
+       "      <td>Heliconius sara</td>\n",
+       "      <td>sara</td>\n",
+       "      <td>Heliconius</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>dorsal</td>\n",
+       "      <td>Madingley</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>tif</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>40112</th>\n",
+       "      <td>26102</td>\n",
+       "      <td>CAM050036</td>\n",
+       "      <td>CAM050036_S1_9_FW_IMG_8547_wb_2.tif</td>\n",
+       "      <td>34adbd5596f977834fb2c6876d2f4493</td>\n",
+       "      <td>2553977</td>\n",
+       "      <td>Heliconius sara ssp. sara</td>\n",
+       "      <td>Heliconius sara</td>\n",
+       "      <td>sara</td>\n",
+       "      <td>Heliconius</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>dorsal</td>\n",
+       "      <td>Gamboa_insectaries</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>tif</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>40116</th>\n",
+       "      <td>26115</td>\n",
+       "      <td>CAM050036</td>\n",
+       "      <td>CAM050036_S1_9_FW_IMG_8547_cut_2.tif</td>\n",
+       "      <td>34adbd5596f977834fb2c6876d2f4493</td>\n",
+       "      <td>2553977</td>\n",
+       "      <td>Heliconius sara ssp. sara</td>\n",
+       "      <td>Heliconius sara</td>\n",
+       "      <td>sara</td>\n",
+       "      <td>Heliconius</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>dorsal</td>\n",
+       "      <td>Gamboa_insectaries</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>tif</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>40138</th>\n",
+       "      <td>26112</td>\n",
+       "      <td>CAM050147</td>\n",
+       "      <td>CAM050147_DS1_IMG_8513_4.tif</td>\n",
+       "      <td>464801449aa127933d9619cb4e6f0a01</td>\n",
+       "      <td>2553977</td>\n",
+       "      <td>Heliconius sara ssp. sara</td>\n",
+       "      <td>Heliconius sara</td>\n",
+       "      <td>sara</td>\n",
+       "      <td>Heliconius</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>dorsal</td>\n",
+       "      <td>Madingley</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>tif</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>40145</th>\n",
+       "      <td>26117</td>\n",
+       "      <td>CAM050147</td>\n",
+       "      <td>CAM050147_DS1_FW_IMG_8513_cut_4.tif</td>\n",
+       "      <td>464801449aa127933d9619cb4e6f0a01</td>\n",
+       "      <td>2553977</td>\n",
+       "      <td>Heliconius sara ssp. sara</td>\n",
+       "      <td>Heliconius sara</td>\n",
+       "      <td>sara</td>\n",
+       "      <td>Heliconius</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>dorsal</td>\n",
+       "      <td>Madingley</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>tif</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>40113</th>\n",
+       "      <td>26105</td>\n",
+       "      <td>CAM050036</td>\n",
+       "      <td>CAM050036_S1_9_HW_wt_IMG_8541_cut3.tif</td>\n",
+       "      <td>6d63d659fbc02369010938bbc229ad4d</td>\n",
+       "      <td>2553977</td>\n",
+       "      <td>Heliconius sara ssp. sara</td>\n",
+       "      <td>Heliconius sara</td>\n",
+       "      <td>sara</td>\n",
+       "      <td>Heliconius</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>dorsal</td>\n",
+       "      <td>Gamboa_insectaries</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>tif</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>40120</th>\n",
+       "      <td>26116</td>\n",
+       "      <td>CAM050036</td>\n",
+       "      <td>CAM050036_S1_9_HW_wt_IMG_8541_cut_3.tif</td>\n",
+       "      <td>6d63d659fbc02369010938bbc229ad4d</td>\n",
+       "      <td>2553977</td>\n",
+       "      <td>Heliconius sara ssp. sara</td>\n",
+       "      <td>Heliconius sara</td>\n",
+       "      <td>sara</td>\n",
+       "      <td>Heliconius</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>dorsal</td>\n",
+       "      <td>Gamboa_insectaries</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>NaN</td>\n",
+       "      <td>tif</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "           X      CAMID                               Image_name  \\\n",
+       "40141  26197  CAM050147      CAM050147_DS1_HW_IMG_8537_cut_3.tif   \n",
+       "40144  26196  CAM050147      CAM050147_DS1_HW_IMG_8537_cut_3.tif   \n",
+       "40112  26102  CAM050036      CAM050036_S1_9_FW_IMG_8547_wb_2.tif   \n",
+       "40116  26115  CAM050036     CAM050036_S1_9_FW_IMG_8547_cut_2.tif   \n",
+       "40138  26112  CAM050147             CAM050147_DS1_IMG_8513_4.tif   \n",
+       "40145  26117  CAM050147      CAM050147_DS1_FW_IMG_8513_cut_4.tif   \n",
+       "40113  26105  CAM050036   CAM050036_S1_9_HW_wt_IMG_8541_cut3.tif   \n",
+       "40120  26116  CAM050036  CAM050036_S1_9_HW_wt_IMG_8541_cut_3.tif   \n",
+       "\n",
+       "                                    md5  record_number  \\\n",
+       "40141  21e726df7076f5c13fe1441126bac4d5        2553977   \n",
+       "40144  21e726df7076f5c13fe1441126bac4d5        2553977   \n",
+       "40112  34adbd5596f977834fb2c6876d2f4493        2553977   \n",
+       "40116  34adbd5596f977834fb2c6876d2f4493        2553977   \n",
+       "40138  464801449aa127933d9619cb4e6f0a01        2553977   \n",
+       "40145  464801449aa127933d9619cb4e6f0a01        2553977   \n",
+       "40113  6d63d659fbc02369010938bbc229ad4d        2553977   \n",
+       "40120  6d63d659fbc02369010938bbc229ad4d        2553977   \n",
+       "\n",
+       "                  Taxonomic_Name          species subspecies       genus  \\\n",
+       "40141  Heliconius sara ssp. sara  Heliconius sara       sara  Heliconius   \n",
+       "40144  Heliconius sara ssp. sara  Heliconius sara       sara  Heliconius   \n",
+       "40112  Heliconius sara ssp. sara  Heliconius sara       sara  Heliconius   \n",
+       "40116  Heliconius sara ssp. sara  Heliconius sara       sara  Heliconius   \n",
+       "40138  Heliconius sara ssp. sara  Heliconius sara       sara  Heliconius   \n",
+       "40145  Heliconius sara ssp. sara  Heliconius sara       sara  Heliconius   \n",
+       "40113  Heliconius sara ssp. sara  Heliconius sara       sara  Heliconius   \n",
+       "40120  Heliconius sara ssp. sara  Heliconius sara       sara  Heliconius   \n",
+       "\n",
+       "      Cross_Type    View            Locality  Sex Dataset file_type  \n",
+       "40141        NaN  dorsal           Madingley  NaN     NaN       tif  \n",
+       "40144        NaN  dorsal           Madingley  NaN     NaN       tif  \n",
+       "40112        NaN  dorsal  Gamboa_insectaries  NaN     NaN       tif  \n",
+       "40116        NaN  dorsal  Gamboa_insectaries  NaN     NaN       tif  \n",
+       "40138        NaN  dorsal           Madingley  NaN     NaN       tif  \n",
+       "40145        NaN  dorsal           Madingley  NaN     NaN       tif  \n",
+       "40113        NaN  dorsal  Gamboa_insectaries  NaN     NaN       tif  \n",
+       "40120        NaN  dorsal  Gamboa_insectaries  NaN     NaN       tif  "
+      ]
+     },
+     "execution_count": 15,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "dupes_2553977 = duplicates.loc[duplicates[\"record_number\"] == 2553977]\n",
+    "dupes_2553977[STATS_COLS].sort_values(\"md5\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "For each of these, we want the one with `_cut_` in the `Image_name`."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 16,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "[26102, 26105, 26112]"
+      ]
+     },
+     "execution_count": 16,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "X_dupes_2553977 = [x for x in list(dupes_2553977[\"X\"]) if \"_cut_\" not in dupes_2553977.loc[dupes_2553977[\"X\"] == x, \"Image_name\"].values[0]]\n",
+    "\n",
+    "# Check we got the ones expected\n",
+    "X_dupes_2553977"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 17,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Add 26196, since both those entries (26196 and 26197) are the same\n",
+    "X_dupes_2553977.append(26196)\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 18,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "X                 36211\n",
+       "CAMID             11968\n",
+       "Image_name        36211\n",
+       "md5               36211\n",
+       "record_number        29\n",
+       "Taxonomic_Name      366\n",
+       "species             246\n",
+       "subspecies          155\n",
+       "genus                94\n",
+       "Cross_Type           30\n",
+       "View                  7\n",
+       "Locality            644\n",
+       "Sex                   3\n",
+       "Dataset               7\n",
+       "file_type             3\n",
+       "dtype: int64"
+      ]
+     },
+     "execution_count": 18,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "df_reduced_5 = df_reduced_4.loc[~df_reduced_4[\"X\"].isin(X_dupes_2553977)]\n",
+    "df_reduced_5[STATS_COLS].nunique()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "4 entries were removed, reducing our number of unique `Image_name`s by 3 (one of the duplicated images had all its metadata duplicated), and everything else has remained consistent. Now the number of unique image names and entries matches the number of unique images. We have a total of 36,211 images, and we will save this reduced CSV as our new master file."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Save Reduced Master File of Unique Images\n",
+    "\n",
+    "Don't include last column indicating duplicate MD5s."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 19,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df_reduced_5[list(df_reduced_5.columns)[:-1]].to_csv(\"../Jiggins_Zenodo_Img_Master.csv\", index = False)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Generate New Master Heliconius and Dorsal Subsets\n",
+    "\n",
+    "## Read in New Master File\n",
+    "\n",
+    "First check on our stats noted in the README."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 20,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "<class 'pandas.core.frame.DataFrame'>\n",
+      "RangeIndex: 36211 entries, 0 to 36210\n",
+      "Data columns (total 31 columns):\n",
+      " #   Column            Non-Null Count  Dtype \n",
+      "---  ------            --------------  ----- \n",
+      " 0   CAMID             36211 non-null  object\n",
+      " 1   X                 36211 non-null  int64 \n",
+      " 2   Image_name        36211 non-null  object\n",
+      " 3   View              35432 non-null  object\n",
+      " 4   zenodo_name       36211 non-null  object\n",
+      " 5   zenodo_link       36211 non-null  object\n",
+      " 6   Sequence          35279 non-null  object\n",
+      " 7   Taxonomic_Name    36211 non-null  object\n",
+      " 8   Locality          23127 non-null  object\n",
+      " 9   Sample_accession  3647 non-null   object\n",
+      " 10  Collected_by      3043 non-null   object\n",
+      " 11  Other_ID          11329 non-null  object\n",
+      " 12  Date              23040 non-null  object\n",
+      " 13  Dataset           29114 non-null  object\n",
+      " 14  Store             27813 non-null  object\n",
+      " 15  Brood             13847 non-null  object\n",
+      " 16  Death_Date        269 non-null    object\n",
+      " 17  Cross_Type        4447 non-null   object\n",
+      " 18  Stage             6 non-null      object\n",
+      " 19  Sex               27514 non-null  object\n",
+      " 20  Unit_Type         25060 non-null  object\n",
+      " 21  file_type         36211 non-null  object\n",
+      " 22  record_number     36211 non-null  int64 \n",
+      " 23  species           36211 non-null  object\n",
+      " 24  subspecies        20400 non-null  object\n",
+      " 25  genus             36211 non-null  object\n",
+      " 26  file_url          36211 non-null  object\n",
+      " 27  hybrid_stat       20948 non-null  object\n",
+      " 28  filename          36211 non-null  object\n",
+      " 29  filepath          36211 non-null  object\n",
+      " 30  md5               36211 non-null  object\n",
+      "dtypes: int64(2), object(29)\n",
+      "memory usage: 8.6+ MB\n"
+     ]
+    }
+   ],
+   "source": [
+    "master = pd.read_csv(\"../Jiggins_Zenodo_Img_Master.csv\", low_memory = False)\n",
+    "master.info()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 21,
+   "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>CAMID</th>\n",
+       "      <th>X</th>\n",
+       "      <th>Image_name</th>\n",
+       "      <th>View</th>\n",
+       "      <th>zenodo_name</th>\n",
+       "      <th>zenodo_link</th>\n",
+       "      <th>Sequence</th>\n",
+       "      <th>Taxonomic_Name</th>\n",
+       "      <th>Locality</th>\n",
+       "      <th>Sample_accession</th>\n",
+       "      <th>...</th>\n",
+       "      <th>file_type</th>\n",
+       "      <th>record_number</th>\n",
+       "      <th>species</th>\n",
+       "      <th>subspecies</th>\n",
+       "      <th>genus</th>\n",
+       "      <th>file_url</th>\n",
+       "      <th>hybrid_stat</th>\n",
+       "      <th>filename</th>\n",
+       "      <th>filepath</th>\n",
+       "      <th>md5</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>13300</th>\n",
+       "      <td>CAM013566</td>\n",
+       "      <td>26040</td>\n",
+       "      <td>13566_H_m_thelxiopeia_V.JPG.jpg</td>\n",
+       "      <td>ventral</td>\n",
+       "      <td>Heliconius_wing_old_photos_2001_2019_part3.csv</td>\n",
+       "      <td>https://zenodo.org/record/2553977</td>\n",
+       "      <td>13,566</td>\n",
+       "      <td>Heliconius melpomene ssp. thelxiopeia</td>\n",
+       "      <td>Maripasoula</td>\n",
+       "      <td>ERS977708</td>\n",
+       "      <td>...</td>\n",
+       "      <td>jpg</td>\n",
+       "      <td>2553977</td>\n",
+       "      <td>Heliconius melpomene</td>\n",
+       "      <td>thelxiopeia</td>\n",
+       "      <td>Heliconius</td>\n",
+       "      <td>https://zenodo.org/record/2553977/files/13566_...</td>\n",
+       "      <td>non-hybrid</td>\n",
+       "      <td>26040_13566_H_m_thelxiopeia_V.JPG.jpg</td>\n",
+       "      <td>images/Heliconius melpomene ssp. thelxiopeia/2...</td>\n",
+       "      <td>b5ed68f937132cbbb04a81d550c9f17e</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>13301</th>\n",
+       "      <td>CAM013566</td>\n",
+       "      <td>26039</td>\n",
+       "      <td>13566_H_m_thelxiopeia_D.JPG.jpg</td>\n",
+       "      <td>dorsal</td>\n",
+       "      <td>Heliconius_wing_old_photos_2001_2019_part3.csv</td>\n",
+       "      <td>https://zenodo.org/record/2553977</td>\n",
+       "      <td>13,566</td>\n",
+       "      <td>Heliconius melpomene ssp. thelxiopeia</td>\n",
+       "      <td>Maripasoula</td>\n",
+       "      <td>ERS977708</td>\n",
+       "      <td>...</td>\n",
+       "      <td>jpg</td>\n",
+       "      <td>2553977</td>\n",
+       "      <td>Heliconius melpomene</td>\n",
+       "      <td>thelxiopeia</td>\n",
+       "      <td>Heliconius</td>\n",
+       "      <td>https://zenodo.org/record/2553977/files/13566_...</td>\n",
+       "      <td>non-hybrid</td>\n",
+       "      <td>26039_13566_H_m_thelxiopeia_D.JPG.jpg</td>\n",
+       "      <td>images/Heliconius melpomene ssp. thelxiopeia/2...</td>\n",
+       "      <td>cb107e2d6a2c3c84b68bbf4e3cfa7af7</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>13302</th>\n",
+       "      <td>CAM013715</td>\n",
+       "      <td>26041</td>\n",
+       "      <td>13715_H_m_meriana_D.JPG.jpg</td>\n",
+       "      <td>dorsal</td>\n",
+       "      <td>Heliconius_wing_old_photos_2001_2019_part3.csv</td>\n",
+       "      <td>https://zenodo.org/record/2553977</td>\n",
+       "      <td>13,715</td>\n",
+       "      <td>Heliconius melpomene ssp. meriana</td>\n",
+       "      <td>N. Wakapou, Maripausala FR GUIANA</td>\n",
+       "      <td>ERS977704</td>\n",
+       "      <td>...</td>\n",
+       "      <td>jpg</td>\n",
+       "      <td>2553977</td>\n",
+       "      <td>Heliconius melpomene</td>\n",
+       "      <td>meriana</td>\n",
+       "      <td>Heliconius</td>\n",
+       "      <td>https://zenodo.org/record/2553977/files/13715_...</td>\n",
+       "      <td>non-hybrid</td>\n",
+       "      <td>26041_13715_H_m_meriana_D.JPG.jpg</td>\n",
+       "      <td>images/Heliconius melpomene ssp. meriana/26041...</td>\n",
+       "      <td>8c77ecb2d8d878f8ac8b097ad361deda</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>13303</th>\n",
+       "      <td>CAM013715</td>\n",
+       "      <td>26042</td>\n",
+       "      <td>13715_H_m_meriana_V.JPG.jpg</td>\n",
+       "      <td>ventral</td>\n",
+       "      <td>Heliconius_wing_old_photos_2001_2019_part3.csv</td>\n",
+       "      <td>https://zenodo.org/record/2553977</td>\n",
+       "      <td>13,715</td>\n",
+       "      <td>Heliconius melpomene ssp. meriana</td>\n",
+       "      <td>N. Wakapou, Maripausala FR GUIANA</td>\n",
+       "      <td>ERS977704</td>\n",
+       "      <td>...</td>\n",
+       "      <td>jpg</td>\n",
+       "      <td>2553977</td>\n",
+       "      <td>Heliconius melpomene</td>\n",
+       "      <td>meriana</td>\n",
+       "      <td>Heliconius</td>\n",
+       "      <td>https://zenodo.org/record/2553977/files/13715_...</td>\n",
+       "      <td>non-hybrid</td>\n",
+       "      <td>26042_13715_H_m_meriana_V.JPG.jpg</td>\n",
+       "      <td>images/Heliconius melpomene ssp. meriana/26042...</td>\n",
+       "      <td>fcf74c73537323bb90127cf92775ac08</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>13304</th>\n",
+       "      <td>CAM013819</td>\n",
+       "      <td>26044</td>\n",
+       "      <td>13819_H_m_meriana_V.JPG.jpg</td>\n",
+       "      <td>ventral</td>\n",
+       "      <td>Heliconius_wing_old_photos_2001_2019_part3.csv</td>\n",
+       "      <td>https://zenodo.org/record/2553977</td>\n",
+       "      <td>13,819</td>\n",
+       "      <td>Heliconius melpomene ssp. meriana</td>\n",
+       "      <td>N. Wakapou, Maripausala FR GUIANA</td>\n",
+       "      <td>ERS977703</td>\n",
+       "      <td>...</td>\n",
+       "      <td>jpg</td>\n",
+       "      <td>2553977</td>\n",
+       "      <td>Heliconius melpomene</td>\n",
+       "      <td>meriana</td>\n",
+       "      <td>Heliconius</td>\n",
+       "      <td>https://zenodo.org/record/2553977/files/13819_...</td>\n",
+       "      <td>non-hybrid</td>\n",
+       "      <td>26044_13819_H_m_meriana_V.JPG.jpg</td>\n",
+       "      <td>images/Heliconius melpomene ssp. meriana/26044...</td>\n",
+       "      <td>601b67512a8c25e5f25a9da89841e8c2</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>13305</th>\n",
+       "      <td>CAM013819</td>\n",
+       "      <td>26043</td>\n",
+       "      <td>13819_H_m_meriana_D.JPG.jpg</td>\n",
+       "      <td>dorsal</td>\n",
+       "      <td>Heliconius_wing_old_photos_2001_2019_part3.csv</td>\n",
+       "      <td>https://zenodo.org/record/2553977</td>\n",
+       "      <td>13,819</td>\n",
+       "      <td>Heliconius melpomene ssp. meriana</td>\n",
+       "      <td>N. Wakapou, Maripausala FR GUIANA</td>\n",
+       "      <td>ERS977703</td>\n",
+       "      <td>...</td>\n",
+       "      <td>jpg</td>\n",
+       "      <td>2553977</td>\n",
+       "      <td>Heliconius melpomene</td>\n",
+       "      <td>meriana</td>\n",
+       "      <td>Heliconius</td>\n",
+       "      <td>https://zenodo.org/record/2553977/files/13819_...</td>\n",
+       "      <td>non-hybrid</td>\n",
+       "      <td>26043_13819_H_m_meriana_D.JPG.jpg</td>\n",
+       "      <td>images/Heliconius melpomene ssp. meriana/26043...</td>\n",
+       "      <td>fe4cf831c5203ad6c9010fc047b7fa10</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "<p>6 rows × 31 columns</p>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "           CAMID      X                       Image_name     View  \\\n",
+       "13300  CAM013566  26040  13566_H_m_thelxiopeia_V.JPG.jpg  ventral   \n",
+       "13301  CAM013566  26039  13566_H_m_thelxiopeia_D.JPG.jpg   dorsal   \n",
+       "13302  CAM013715  26041      13715_H_m_meriana_D.JPG.jpg   dorsal   \n",
+       "13303  CAM013715  26042      13715_H_m_meriana_V.JPG.jpg  ventral   \n",
+       "13304  CAM013819  26044      13819_H_m_meriana_V.JPG.jpg  ventral   \n",
+       "13305  CAM013819  26043      13819_H_m_meriana_D.JPG.jpg   dorsal   \n",
+       "\n",
+       "                                          zenodo_name  \\\n",
+       "13300  Heliconius_wing_old_photos_2001_2019_part3.csv   \n",
+       "13301  Heliconius_wing_old_photos_2001_2019_part3.csv   \n",
+       "13302  Heliconius_wing_old_photos_2001_2019_part3.csv   \n",
+       "13303  Heliconius_wing_old_photos_2001_2019_part3.csv   \n",
+       "13304  Heliconius_wing_old_photos_2001_2019_part3.csv   \n",
+       "13305  Heliconius_wing_old_photos_2001_2019_part3.csv   \n",
+       "\n",
+       "                             zenodo_link Sequence  \\\n",
+       "13300  https://zenodo.org/record/2553977   13,566   \n",
+       "13301  https://zenodo.org/record/2553977   13,566   \n",
+       "13302  https://zenodo.org/record/2553977   13,715   \n",
+       "13303  https://zenodo.org/record/2553977   13,715   \n",
+       "13304  https://zenodo.org/record/2553977   13,819   \n",
+       "13305  https://zenodo.org/record/2553977   13,819   \n",
+       "\n",
+       "                              Taxonomic_Name  \\\n",
+       "13300  Heliconius melpomene ssp. thelxiopeia   \n",
+       "13301  Heliconius melpomene ssp. thelxiopeia   \n",
+       "13302      Heliconius melpomene ssp. meriana   \n",
+       "13303      Heliconius melpomene ssp. meriana   \n",
+       "13304      Heliconius melpomene ssp. meriana   \n",
+       "13305      Heliconius melpomene ssp. meriana   \n",
+       "\n",
+       "                                Locality Sample_accession  ... file_type  \\\n",
+       "13300                        Maripasoula        ERS977708  ...       jpg   \n",
+       "13301                        Maripasoula        ERS977708  ...       jpg   \n",
+       "13302  N. Wakapou, Maripausala FR GUIANA        ERS977704  ...       jpg   \n",
+       "13303  N. Wakapou, Maripausala FR GUIANA        ERS977704  ...       jpg   \n",
+       "13304  N. Wakapou, Maripausala FR GUIANA        ERS977703  ...       jpg   \n",
+       "13305  N. Wakapou, Maripausala FR GUIANA        ERS977703  ...       jpg   \n",
+       "\n",
+       "      record_number               species   subspecies       genus  \\\n",
+       "13300       2553977  Heliconius melpomene  thelxiopeia  Heliconius   \n",
+       "13301       2553977  Heliconius melpomene  thelxiopeia  Heliconius   \n",
+       "13302       2553977  Heliconius melpomene      meriana  Heliconius   \n",
+       "13303       2553977  Heliconius melpomene      meriana  Heliconius   \n",
+       "13304       2553977  Heliconius melpomene      meriana  Heliconius   \n",
+       "13305       2553977  Heliconius melpomene      meriana  Heliconius   \n",
+       "\n",
+       "                                                file_url hybrid_stat  \\\n",
+       "13300  https://zenodo.org/record/2553977/files/13566_...  non-hybrid   \n",
+       "13301  https://zenodo.org/record/2553977/files/13566_...  non-hybrid   \n",
+       "13302  https://zenodo.org/record/2553977/files/13715_...  non-hybrid   \n",
+       "13303  https://zenodo.org/record/2553977/files/13715_...  non-hybrid   \n",
+       "13304  https://zenodo.org/record/2553977/files/13819_...  non-hybrid   \n",
+       "13305  https://zenodo.org/record/2553977/files/13819_...  non-hybrid   \n",
+       "\n",
+       "                                    filename  \\\n",
+       "13300  26040_13566_H_m_thelxiopeia_V.JPG.jpg   \n",
+       "13301  26039_13566_H_m_thelxiopeia_D.JPG.jpg   \n",
+       "13302      26041_13715_H_m_meriana_D.JPG.jpg   \n",
+       "13303      26042_13715_H_m_meriana_V.JPG.jpg   \n",
+       "13304      26044_13819_H_m_meriana_V.JPG.jpg   \n",
+       "13305      26043_13819_H_m_meriana_D.JPG.jpg   \n",
+       "\n",
+       "                                                filepath  \\\n",
+       "13300  images/Heliconius melpomene ssp. thelxiopeia/2...   \n",
+       "13301  images/Heliconius melpomene ssp. thelxiopeia/2...   \n",
+       "13302  images/Heliconius melpomene ssp. meriana/26041...   \n",
+       "13303  images/Heliconius melpomene ssp. meriana/26042...   \n",
+       "13304  images/Heliconius melpomene ssp. meriana/26044...   \n",
+       "13305  images/Heliconius melpomene ssp. meriana/26043...   \n",
+       "\n",
+       "                                    md5  \n",
+       "13300  b5ed68f937132cbbb04a81d550c9f17e  \n",
+       "13301  cb107e2d6a2c3c84b68bbf4e3cfa7af7  \n",
+       "13302  8c77ecb2d8d878f8ac8b097ad361deda  \n",
+       "13303  fcf74c73537323bb90127cf92775ac08  \n",
+       "13304  601b67512a8c25e5f25a9da89841e8c2  \n",
+       "13305  fe4cf831c5203ad6c9010fc047b7fa10  \n",
+       "\n",
+       "[6 rows x 31 columns]"
+      ]
+     },
+     "execution_count": 21,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "master.loc[master.Stage.notna()]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 22,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "CAMID                820\n",
+       "X                   4447\n",
+       "Image_name          4447\n",
+       "View                   2\n",
+       "zenodo_name            4\n",
+       "zenodo_link            4\n",
+       "Sequence             820\n",
+       "Taxonomic_Name         8\n",
+       "Locality               0\n",
+       "Sample_accession       0\n",
+       "Collected_by           0\n",
+       "Other_ID               0\n",
+       "Date                 160\n",
+       "Dataset                1\n",
+       "Store                 10\n",
+       "Brood                 49\n",
+       "Death_Date             0\n",
+       "Cross_Type            30\n",
+       "Stage                  0\n",
+       "Sex                    2\n",
+       "Unit_Type              1\n",
+       "file_type              2\n",
+       "record_number          4\n",
+       "species                2\n",
+       "subspecies            21\n",
+       "genus                  1\n",
+       "file_url            4447\n",
+       "hybrid_stat            2\n",
+       "filename            4447\n",
+       "filepath            4447\n",
+       "md5                 4447\n",
+       "dtype: int64"
+      ]
+     },
+     "execution_count": 22,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "master.loc[master.Cross_Type.notna()].nunique()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 23,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "View\n",
+       "dorsal                17235\n",
+       "ventral               17167\n",
+       "hindwing dorsal         254\n",
+       "hindwing ventral        254\n",
+       "forewing dorsal         252\n",
+       "forewing ventral        252\n",
+       "dorsal and ventral       18\n",
+       "Name: count, dtype: int64"
+      ]
+     },
+     "execution_count": 23,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "master.View.value_counts()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Generate master_licenses CSV file to re-run [`get_licenses.py`](https://huggingface.co/datasets/imageomics/Comparison-Subset-Jiggins/blob/main/scripts/get_licenses.py) for nicely formatted licensing info of the 29 included records.\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 24,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "(29, 2)"
+      ]
+     },
+     "execution_count": 24,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "master_licenses = master[[\"record_number\", \"zenodo_link\"]]\n",
+    "master_licenses = master_licenses.loc[~master_licenses.duplicated([\"record_number\", \"zenodo_link\"], keep = \"first\")]\n",
+    "master_licenses.shape"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 25,
+   "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>record_number</th>\n",
+       "      <th>url</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>0</th>\n",
+       "      <td>4289223</td>\n",
+       "      <td>https://zenodo.org/record/4289223</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>145</th>\n",
+       "      <td>4288311</td>\n",
+       "      <td>https://zenodo.org/record/4288311</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "     record_number                                url\n",
+       "0          4289223  https://zenodo.org/record/4289223\n",
+       "145        4288311  https://zenodo.org/record/4288311"
+      ]
+     },
+     "execution_count": 25,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "master_licenses.rename(columns = {\"zenodo_link\": \"url\"}, inplace = True)\n",
+    "master_licenses.head(2)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 26,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Save to CSV to run through get licenses script\n",
+    "master_licenses.to_csv(\"../metadata/master_licenses.csv\", index = False)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Make Heliconius Master Subset\n",
+    "\n",
+    "Now make the Heliconius master CSV, and check stats to update README."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 27,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "CAMID               10086\n",
+       "X                   29134\n",
+       "Image_name          29134\n",
+       "View                    3\n",
+       "zenodo_name            31\n",
+       "zenodo_link            29\n",
+       "Sequence             9008\n",
+       "Taxonomic_Name        132\n",
+       "Locality              471\n",
+       "Sample_accession     1559\n",
+       "Collected_by           12\n",
+       "Other_ID             1865\n",
+       "Date                  773\n",
+       "Dataset                 7\n",
+       "Store                 121\n",
+       "Brood                 224\n",
+       "Death_Date             79\n",
+       "Cross_Type             30\n",
+       "Stage                   1\n",
+       "Sex                     3\n",
+       "Unit_Type               4\n",
+       "file_type               3\n",
+       "record_number          29\n",
+       "species                37\n",
+       "subspecies            110\n",
+       "genus                   1\n",
+       "file_url            29134\n",
+       "hybrid_stat             2\n",
+       "filename            29134\n",
+       "filepath            29134\n",
+       "md5                 29134\n",
+       "dtype: int64"
+      ]
+     },
+     "execution_count": 27,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "heliconius_master = master.loc[master.genus.str.lower() == \"heliconius\"].copy()\n",
+    "heliconius_master.nunique()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 28,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "View\n",
+       "dorsal                14202\n",
+       "ventral               14135\n",
+       "dorsal and ventral       18\n",
+       "Name: count, dtype: int64"
+      ]
+     },
+     "execution_count": 28,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "heliconius_master.View.value_counts()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "### Save Heliconius Subset to CSV"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 29,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "heliconius_master.to_csv(\"../Jiggins_Heliconius_Master.csv\", index = False)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Make Dorsal Subset\n",
+    "\n",
+    "We'll now make a CSV of all dorsal images."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 30,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "CAMID               11752\n",
+       "X                   17759\n",
+       "Image_name          17759\n",
+       "View                    4\n",
+       "zenodo_name            31\n",
+       "zenodo_link            29\n",
+       "Sequence            10689\n",
+       "Taxonomic_Name        362\n",
+       "Locality              641\n",
+       "Sample_accession     1552\n",
+       "Collected_by           12\n",
+       "Other_ID             2890\n",
+       "Date                  788\n",
+       "Dataset                 7\n",
+       "Store                 137\n",
+       "Brood                 215\n",
+       "Death_Date             61\n",
+       "Cross_Type             30\n",
+       "Stage                   1\n",
+       "Sex                     3\n",
+       "Unit_Type               4\n",
+       "file_type               3\n",
+       "record_number          29\n",
+       "species               245\n",
+       "subspecies            152\n",
+       "genus                  94\n",
+       "file_url            17759\n",
+       "hybrid_stat             2\n",
+       "filename            17759\n",
+       "filepath            17759\n",
+       "md5                 17759\n",
+       "dtype: int64"
+      ]
+     },
+     "execution_count": 30,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "dorsal_views = [view for view in list(master.View.dropna().unique()) if \"dorsal\" in view]\n",
+    "\n",
+    "dorsal_master = master.loc[master[\"View\"].isin(dorsal_views)].copy()\n",
+    "dorsal_master.nunique()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 31,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "CAM_dupe\n",
+       "False    15460\n",
+       "True      2299\n",
+       "Name: count, dtype: int64"
+      ]
+     },
+     "execution_count": 31,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "dorsal_master[\"CAM_dupe\"] = dorsal_master.duplicated([\"CAMID\", \"file_type\"], keep = False)\n",
+    "dorsal_master.CAM_dupe.value_counts()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 32,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "CAM_dupe\n",
+       "False          15442\n",
+       "True            1793\n",
+       "single-wing      506\n",
+       "both-wings        18\n",
+       "Name: count, dtype: int64"
+      ]
+     },
+     "execution_count": 32,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# Adding label for both wings alters counts, as in, some only have the double wing image\n",
+    "dorsal_master.loc[dorsal_master[\"View\"].isin([\"forewing dorsal\", \"hindwing dorsal\"]), \"CAM_dupe\"] = \"single-wing\"\n",
+    "dorsal_master.loc[dorsal_master[\"View\"] == \"dorsal and ventral\", \"CAM_dupe\"] = \"both-wings\"\n",
+    "dorsal_master.CAM_dupe.value_counts()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 33,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "881\n",
+      "11452\n"
+     ]
+    },
+    {
+     "data": {
+      "text/plain": [
+       "13"
+      ]
+     },
+     "execution_count": 33,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "print(dorsal_master.loc[dorsal_master[\"CAM_dupe\"] == True, \"CAMID\"].nunique())\n",
+    "print(dorsal_master.loc[dorsal_master[\"CAM_dupe\"] == False, \"CAMID\"].nunique())\n",
+    "dorsal_master.loc[dorsal_master[\"CAM_dupe\"] == True, \"record_number\"].nunique()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 34,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "file_type\n",
+       "raw    253\n",
+       "jpg    253\n",
+       "Name: count, dtype: int64"
+      ]
+     },
+     "execution_count": 34,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "dorsal_master.loc[dorsal_master[\"CAM_dupe\"] == \"single-wing\", \"file_type\"].value_counts()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 35,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "file_type\n",
+       "jpg    17\n",
+       "tif     1\n",
+       "Name: count, dtype: int64"
+      ]
+     },
+     "execution_count": 35,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "dorsal_master.loc[dorsal_master[\"CAM_dupe\"] == \"both-wings\", \"file_type\"].value_counts()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 36,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "file_type\n",
+       "jpg    1775\n",
+       "tif      16\n",
+       "raw       2\n",
+       "Name: count, dtype: int64"
+      ]
+     },
+     "execution_count": 36,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "dorsal_master.loc[dorsal_master[\"CAM_dupe\"] == True, \"file_type\"].value_counts()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 37,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "file_type\n",
+       "jpg    10706\n",
+       "raw     4718\n",
+       "tif       18\n",
+       "Name: count, dtype: int64"
+      ]
+     },
+     "execution_count": 37,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "dorsal_master.loc[dorsal_master[\"CAM_dupe\"] == False, \"file_type\"].value_counts()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 38,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "False    10723\n",
+       "True      2028\n",
+       "Name: count, dtype: int64"
+      ]
+     },
+     "execution_count": 38,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "dorsal_master.loc[dorsal_master[\"file_type\"] == \"jpg\"].duplicated(subset = \"CAMID\", keep = False).value_counts()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 39,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "False    4719\n",
+       "True      254\n",
+       "Name: count, dtype: int64"
+      ]
+     },
+     "execution_count": 39,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "dorsal_master.loc[dorsal_master[\"file_type\"] == \"raw\"].duplicated(subset = \"CAMID\", keep = False).value_counts()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Generally speaking, it seems RAW images will be unique, as there is only one CAMID that gets duplicated across the raw images (2 photos of that specimen that are RAW) outside of the single wing views.\n",
+    "\n",
+    "### Save Dorsal Subset to CSV"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 40,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "dorsal_master.to_csv(\"../Jiggins_Zenodo_dorsal_Img_Master.csv\", index = False)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "std",
+   "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": "3.11.3"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}