2018-05-16-library-catalog-subject.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Plotting Library Catalog Subjects\n",
    "\n",
    "Kaggle has a cool dataset of the Seattle Public Library's catalog. Each item has a list of subjects. For example, the first entry has:\n",
    "\n",
    "    Musicians Fiction, Bullfighters Fiction, Best friends Fiction, Friendship Fiction, Adventure and adventurers Fiction\n",
    "\n",
    "Using what I learned in my NLP course last semester, I used one method of building word vectors to make vectors representing library catalog subjects. When I plot them I get something like this:\n",
    "\n",
    "![](images/2018-05-16-labeled-plot.png)\n",
    "\n",
    "### Motivation\n",
    "\n",
    "Natural language processing has techniques for converting words into continuous vector representations: `fish` can be represented with a vector like `[0.2, -0.23, 0.0]`. These sometimes contain interesting information (e.g. [\"king - man + woman = queen\"](http://p.migdal.pl/2017/01/06/king-man-woman-queen-why.html)). \n",
    "\n",
    "I can use one of those techniques to come up with vectors representing library catalog subjects.\n",
    "\n",
    "Primarily I wanted to use the vectors to plot the library book subjects. I'm curious if I'd see a dimension representing the \"children's books - young adult - grown up\", or if science subjects cluster together.\n",
    "\n",
    "Word vectors are also useful in Information Retrieval. With a vector representation, I can use math to measure how similar two subjects are. For example, if I'm searching the catalog for \"Friendship Fiction\", I wouldn't mind also seeing \"Best friends Fiction\" books that are missing the \"Friendship Fiction\" label. Subject vectors could help.\n",
    "\n",
    "Word vectors are also used in more complex NLP models like neural machine translation.\n",
    "The way I think of this is that if I can tell the model that \"Best friends Fiction\" and \"Friendship Fiction\" are related in a specific way, then when the model learns about \"Best friends Fiction\", it can also improve predictions for \"Friendship Fiction.\"\n",
    "\n",
    "### Vector representations\n",
    "\n",
    "In NLP-land, \"distributional methods\" assume that \"Words that occur in similar contexts tend to have similar meanings\" (from [Jurafsky and Martin's textbook](http://web.stanford.edu/~jurafsky/slp3/6.pdf)). I think distributional methods would make interesting representations of library catalog subjects.\n",
    "One way to turn this into word vectors is to make sparse vectors of metrics based on co-occurring words, and then reduce the number of dimensions to get a dense vector.\n",
    "\n",
    "\n",
    "### Plan\n",
    "\n",
    "Using this assumption with library catalog subjects, I can describe \"Musicians Fiction\" with the subjects \"Bullfighters Fiction\", \"Best friends Fiction\", and so on. Specifically, I'll go through all catalog entries that contain \"Musicians Fiction\" and count up how many times each other subject shows up. This will give me a huge vector of counts like:\n",
    "\n",
    "     Musicians Fiction:   0 1 4 0 ... 150K other numbers, mostly 0s ... 1 0 1 0\n",
    "\n",
    "I could store this in a big `Number of subjects x Number of subjects` matrix, but it turns out there are a lot of subjects (~150K)! Because of computational limitations, I'll define a subset of subjects containing the \"interesting\" subjects (~1000), and only compute dense vectors for these. This means I'll use a matrix of size `Number of interesting subjects x Number of subjects` (1000 x 150K). By doing it this way, I'll still use the rarer subjects when coming up with the dense vectors of the interesting subjects.\n",
    "\n",
    "Raw co-occurrence counts and relative frequencies have some issues. For word vectors, one problem is that words like \"the\" occur with all subjects equally (or for library catalog subject vectors, maybe \"Large print books\" co-occur with many other subjects). These subjects don't provide more information about the interesting subjects. So instead of using the counts directly, I'll use PPMI, which is higher if the words co-occur more often than chance. It is a function of the co-occurrence counts and global counts.\n",
    "\n",
    "At this point, I'll have sparse vectors of length `Number of subjects` (~150K) that represent each of my interesting subjects. I'll use PCA to reduce the number of dimensions. Dimensionality reduction is really cool! By reducing the number of dimensions in a clever way, the vector might represent higher-order co-occurrence. [Here's a post on reducing dimensions using SVD](2018-05-16-SVD-demo.ipynb).\n",
    "\n",
    "This was a quick introduction to dense vectors. Check out [Vector Semantics](https://web.stanford.edu/~jurafsky/slp3/6.pdf) for details!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# imports and helper functions you can skip over :D\n",
    "import csv\n",
    "from collections import defaultdict\n",
    "import os\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import scipy.sparse as sp\n",
    "from scipy.sparse.linalg import svds\n",
    "\n",
    "SAVE = True\n",
    "def maybe_save_plot(filename):\n",
    "    if SAVE:\n",
    "        plt.tight_layout()\n",
    "        plt.savefig('images/' + filename, bbox_inches=\"tight\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data\n",
    "\n",
    "I got this data from [Kaggle](http://kaggle.com/seattle-public-library), though it's also available at [data.seattle.gov](https://data.seattle.gov/Community/Checkouts-by-Title/tmmm-ytt6). I have it downloaded and unzipped locally in the `data` folder. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "LOCATION = 'data/seattle-library-checkout-records/'\n",
    "file_path = os.path.join(LOCATION, 'Library_Collection_Inventory.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here's what a line of the `Library_Collection_Inventory.csv` file contains:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('BibNum', '3011076')\n",
      "('Title', \"A tale of two friends / adapted by Ellie O'Ryan ; illustrated by Tom Caulfield, Frederick Gardner, Megan Petasky, and Allen Tam.\")\n",
      "('Author', \"O'Ryan, Ellie\")\n",
      "('ISBN', '1481425730, 1481425749, 9781481425735, 9781481425742')\n",
      "('PublicationYear', '2014.')\n",
      "('Publisher', 'Simon Spotlight,')\n",
      "('Subjects', 'Musicians Fiction, Bullfighters Fiction, Best friends Fiction, Friendship Fiction, Adventure and adventurers Fiction')\n",
      "('ItemType', 'jcbk')\n",
      "('ItemCollection', 'ncrdr')\n",
      "('FloatingItem', 'Floating')\n",
      "('ItemLocation', 'qna')\n",
      "('ReportDate', '09/01/2017')\n",
      "('ItemCount', '1')\n"
     ]
    }
   ],
   "source": [
    "# print the first entry\n",
    "with open(file_path, 'r') as csvfile:\n",
    "    reader = csv.DictReader(csvfile)\n",
    "    print('\\n'.join(map(str, next(reader).items())))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "    ('BibNum', '3011076')\n",
    "    ('Title', \"A tale of two friends / adapted by Ellie O'Ryan ; illustrated by Tom Caulfield, Frederick Gardner, Megan Petasky, and Allen Tam.\")\n",
    "    ('Author', \"O'Ryan, Ellie\")\n",
    "    ('ISBN', '1481425730, 1481425749, 9781481425735, 9781481425742')\n",
    "    ('PublicationYear', '2014.')\n",
    "    ('Publisher', 'Simon Spotlight,')\n",
    "    ('Subjects', 'Musicians Fiction, Bullfighters Fiction, Best friends Fiction, Friendship Fiction, Adventure and adventurers Fiction')\n",
    "    ('ItemType', 'jcbk')\n",
    "    ('ItemCollection', 'ncrdr')\n",
    "    ('FloatingItem', 'Floating')\n",
    "    ('ItemLocation', 'qna')\n",
    "    ('ReportDate', '09/01/2017')\n",
    "    ('ItemCount', '1')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Preprocessing\n",
    "\n",
    "I'll start by getting the list of subjects out of the file and perform a little preprocessing.\n",
    "\n",
    "First, I'll drop items without subjects. Then I'll work with lowercase strings.\n",
    "\n",
    "Then I'll remove duplicates. If I have duplicate items with the same subjects, I'll overestimate how closely two subjects are related. A complication is that the `Title` field alone isn't enough to detect duplicates. For example, here are a few `Title` fields for \"Crime and Punishment\":\n",
    "\n",
    "    Crime and punishment / Fyodor Dostoyevsky ; translated with an introduction and notes by Oliver Ready.\n",
    "    Crime and punishment / Fyodor Dostoevsky ; translated from the Russian by Richard Pevear and Larissa Volokhonsky\n",
    "    Crime and punishment / Fyodor Dostoyevsky ; translated with an introduction and notes by David McDuff.\n",
    "    Crime and punishment / Fyodor Dostoyevsky ; translated with an introduction and notes by David McDuff.\n",
    "    Crime and punishment / Fyodor Dostoyevsky ; translated with an introduction and notes by David McDuff.\n",
    "        \n",
    "A naive way to mitigate this is to count an item as unique if it has the same title (the text before the `/` in the `Title` field) and author (from the `Author` field), and take the union of all duplicate entries' subjects."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "items_to_subjects = defaultdict(set)\n",
    "\n",
    "def clean_text(text):\n",
    "    return text.strip().lower()\n",
    "\n",
    "def normalize_item(item):\n",
    "    # TODO: The library system has multiple copies of the same book. \n",
    "    # This catches a few books with exactly the same title/author, because\n",
    "    # otherwise I'll overcount their subjects' co-occurrences. But this approach\n",
    "    # can be improved!\n",
    "    title = item['Title'].split('/')[0]  # The stuff before '/' looks like the title\n",
    "    return (\n",
    "        clean_text(title),\n",
    "        clean_text(item['Author']),\n",
    "    )\n",
    "\n",
    "with open(file_path, 'r') as csvfile:\n",
    "    reader = csv.DictReader(csvfile)\n",
    "    for i, item in enumerate(reader):\n",
    "        subjects = [\n",
    "            clean_text(item)\n",
    "            for item in item['Subjects'].split(', ')\n",
    "        ]\n",
    "        \n",
    "        # Drop items without subjects!\n",
    "        if not subjects:\n",
    "            continue\n",
    "            \n",
    "        # normalize the item\n",
    "        item_info = normalize_item(item)\n",
    "        \n",
    "        items_to_subjects[item_info].update(subjects)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Pandas\n",
    "\n",
    "Now I'll expand `items_to_subjects` into a list of `item-subject` pairs and make a `pandas` DataFrame from it. I just use `pandas` to make it a little easier to compute counts."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_item_subject_pairs_df = pd.DataFrame(\n",
    "    [\n",
    "        (item, subject)\n",
    "        for item, subjects in items_to_subjects.items()\n",
    "        for subject in subjects\n",
    "    ], \n",
    "    columns=['Item', 'Subject']\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style>\n",
       "    .dataframe thead tr:only-child th {\n",
       "        text-align: right;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Item</th>\n",
       "      <th>Subject</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>(a tale of two friends, o'ryan, ellie)</td>\n",
       "      <td>musicians fiction</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>(a tale of two friends, o'ryan, ellie)</td>\n",
       "      <td>bullfighters fiction</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>(a tale of two friends, o'ryan, ellie)</td>\n",
       "      <td>best friends fiction</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>(a tale of two friends, o'ryan, ellie)</td>\n",
       "      <td>friendship fiction</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>(a tale of two friends, o'ryan, ellie)</td>\n",
       "      <td>adventure and adventurers fiction</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                     Item                            Subject\n",
       "0  (a tale of two friends, o'ryan, ellie)                  musicians fiction\n",
       "1  (a tale of two friends, o'ryan, ellie)               bullfighters fiction\n",
       "2  (a tale of two friends, o'ryan, ellie)               best friends fiction\n",
       "3  (a tale of two friends, o'ryan, ellie)                 friendship fiction\n",
       "4  (a tale of two friends, o'ryan, ellie)  adventure and adventurers fiction"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "all_item_subject_pairs_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Stats\n",
    "\n",
    "Now I can check out a few stats. There are around 450K subjects. 303K subjects occur for only one book, including `bridges northwest pacific juvenile fiction` and `cicindela speciation`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total items: 1817414\n",
      "total unique subjects: 450371\n",
      "total unique items: 563868\n",
      "\n",
      "subjects with one book: 303239\n"
     ]
    }
   ],
   "source": [
    "print('total items:', len(all_item_subject_pairs_df))\n",
    "print('total unique subjects:', len(all_item_subject_pairs_df.groupby('Subject')))\n",
    "print('total unique items:', len(all_item_subject_pairs_df.groupby('Item')))\n",
    "print()\n",
    "print('subjects with one book:', sum(all_item_subject_pairs_df['Subject'].value_counts() == 1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Removing subjects with only one item\n",
    "\n",
    "Subjects that only occur once aren't going to give much information about other subjects, so I'll filter them out.\n",
    "This gives me around 150K subjects."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "item_subject_pairs_df = all_item_subject_pairs_df.groupby('Subject').filter(lambda x: len(x) > 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total unique subjects: 147132\n",
      "subjects with one book (should be 0): 0\n"
     ]
    }
   ],
   "source": [
    "print('total unique subjects:', len(item_subject_pairs_df.groupby('Subject')))\n",
    "print('subjects with one book (should be 0):', sum(item_subject_pairs_df['Subject'].value_counts() == 1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Mappings"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Subject to index mappings\n",
    "\n",
    "First I need to map each subject to an index for the `Number of subjects` dimension. To do this I'll have a way to convert an index into a subject and a subject into an index."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "index_to_other_subject = item_subject_pairs_df['Subject'].unique()\n",
    "\n",
    "other_subject_to_index = {\n",
    "    subject: i\n",
    "    for i, subject in enumerate(index_to_other_subject)\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### \"Interesting\" subjects\n",
    "\n",
    "To make things run a little faster, I'll also define a set of subjects as the ones I'll want to plot.\n",
    "I'll use subjects that occur at least 100 times."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "INTERESTING_SUBJECT_COUNT_CUTOFF = 100\n",
    "\n",
    "index_to_interesting_subject = item_subject_pairs_df.groupby('Subject').filter(\n",
    "    lambda x: len(x) > INTERESTING_SUBJECT_COUNT_CUTOFF\n",
    ")['Subject'].unique()\n",
    "\n",
    "interesting_subject_to_index = {\n",
    "    subject: i\n",
    "    for i, subject in enumerate(index_to_interesting_subject)\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Subject counts\n",
    "\n",
    "To compute PPMI, I'll also need the number of overall subject counts."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "subject_to_count = dict(item_subject_pairs_df['Subject'].value_counts())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Sparse Matrix\n",
    "\n",
    "\n",
    "\n",
    "Now I want to load this into a subject-subject matrix of size the `Number of _interesting_ subjects x Number of subjects` matrix. Each row will represent the \"interesting\" subjects and columns that represent information I know about those subjects, in this case, the PPMI on co-occurring subjects.\n",
    "\n",
    "PPMI is given by \n",
    "\n",
    "$$PPMI(w_i, w_j) =  \\max\\left(\\log\\left( \\frac{n \\,\\times\\, \\text{count of co-occurrences of }w_i\\text{ and }w_j}{\\text{count of } w_i  \\, \\times \\, \\text{count of } w_j}\\right), 0\\right).$$\n",
    "\n",
    "If you're interested in more information about PPMI and what I'm up to, \n",
    "check out Vector Semantics and Semantics with Dense Vectors in [Speech and Language Processing](http://web.stanford.edu/~jurafsky/slp3/). Specifically: \"Weighing terms: Pointwise Mutual Information (PMI)\", currently in section 15.2.\n",
    "\n",
    "Because this matrix would have 255,000,000 elements, I use `scipy`'s sparse matrices. I use the `coo_matrix`, which is supposed to be good for constructing sparse matrices, and then convert it to a `csr_matrix`, which supports arithmetic operations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def word_word_sparse_from_data(\n",
    "    item_to_subjects,\n",
    "    subject_to_count, \n",
    "    interesting_subject_to_index,\n",
    "    other_subject_to_index,\n",
    "):\n",
    "    '''Build a sparse matrix of the subject-subject co-occurrence PPMI's\n",
    "    \n",
    "    returns:  a sparse matrix with size `Number of interesting subjects` x `Number of subjects`\n",
    "    '''\n",
    "    # First count how many times subjects d\n",
    "    # This builds a dictionary that maps an (interesting_subject, other_subject) to a count of co-occurrences.\n",
    "    # The interesting_subject should be in the interesting_subject list defined above (i.e. filter out non-interesting)\n",
    "    # The other_subject should be in the (almost) full list of words (i.e. remove words we filtered from other_subject_to_index)\n",
    "    counts = defaultdict(int)\n",
    "    for title, subjects in item_to_subjects.items():\n",
    "        for interesting_subject in subjects:\n",
    "            # the interesting subject should be in the interesting subjects list\n",
    "            if interesting_subject not in interesting_subject_to_index:\n",
    "                continue\n",
    "            for other_subject in subjects:\n",
    "                # The other subject should not be the same as the main one\n",
    "                if interesting_subject == other_subject:\n",
    "                    continue                \n",
    "                # The other subject needs to be in the larger subject list (occurring at least once)\n",
    "                if other_subject not in other_subject_to_index:\n",
    "                    continue\n",
    "                counts[(interesting_subject, other_subject)] += 1\n",
    "\n",
    "    total = sum(subject_to_count.values())\n",
    "     \n",
    "    # build the three lists I need for the `sp.coo_matrix` call.\n",
    "    interesting_subject_ids = []\n",
    "    other_subject_ids = []\n",
    "    count_ids = []\n",
    "        \n",
    "    for (interesting_subject, other_subject), count in counts.items():\n",
    "        c_i = subject_to_count[interesting_subject]\n",
    "        c_j = subject_to_count[other_subject]\n",
    "        \n",
    "        # PPMI\n",
    "        new_count = np.max(np.log(total * count/(c_i * c_j)), 0)\n",
    "        \n",
    "        # Find these subjects' indices\n",
    "        interesting_subject_ids.append(interesting_subject_to_index[interesting_subject])\n",
    "        other_subject_ids.append(other_subject_to_index[other_subject])\n",
    "        count_ids.append(new_count)\n",
    "    \n",
    "    # Build a sparse matrix from these lists\n",
    "    sparse_word_word = sp.coo_matrix((\n",
    "        count_ids, \n",
    "        (interesting_subject_ids, other_subject_ids)\n",
    "    )).astype(float)\n",
    "    return sparse_word_word.tocsr()\n",
    "\n",
    "sparse_word_word = word_word_sparse_from_data(\n",
    "    items_to_subjects, \n",
    "    subject_to_count, \n",
    "    interesting_subject_to_index,\n",
    "    other_subject_to_index,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Looking at `sparse_word_word`\n",
    "\n",
    "Using this I could see whether  `political fiction` and `satirical literature` co-occur more or less likely than you'd expect according to its PPMI. In this case, it's `5.281`, so they occur together pretty often!\n",
    "\n",
    "Compare this with the PPMI of `political fiction` and `alphabet` books. This gets a 0 because they co-occurred less frequently than you'd expect."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PPMI for `political fiction` and `satirical literature`: 5.281\n",
      "PPMI for `political fiction` and `alphabet`: 0.000\n"
     ]
    }
   ],
   "source": [
    "def ppmi_for_subjs(subj1, subj2):\n",
    "    print('PPMI for `{}` and `{}`: {:.3f}'.format(\n",
    "        subj1,\n",
    "        subj2,\n",
    "        sparse_word_word[interesting_subject_to_index[subj1], other_subject_to_index[subj2]]\n",
    "    ))\n",
    "\n",
    "ppmi_for_subjs('political fiction', 'satirical literature')\n",
    "ppmi_for_subjs('political fiction', 'alphabet')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Reducing dimensions\n",
    "\n",
    "Now I can use SVD to reduce the number of dimensions. Since I'm just going to plot the points, I'll just reduce it to four dimensions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def dense_vecs_from_sparse(sparse_matrix, dimensions):\n",
    "    # center the matrix\n",
    "    sparse_matrix = sparse_matrix - np.mean(sparse_matrix, axis=0)\n",
    "    u, s, v = svds(sparse_matrix, k=dimensions)\n",
    "\n",
    "    return u"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "dimensions = 4\n",
    "\n",
    "vecs = dense_vecs_from_sparse(\n",
    "    sparse_word_word, \n",
    "    dimensions=dimensions, \n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualizing\n",
    "\n",
    "Let's try to visualize these points in 4-space! \n",
    "I'll draw pairwise plots for each pair of dimensions. I'll also mark the same point, in this case 'documentary film', in all graphs."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_in_4d(axes, four_dim_vec, plot_args=None, plot_kwargs=None, label=''):\n",
    "    plot_args = plot_args or ['.']\n",
    "    plot_kwargs = plot_kwargs or {}\n",
    "    for i in range(3):\n",
    "        for j in range(i + 1, 4):\n",
    "            axes[j - 1, i].plot(four_dim_vec[:, i], four_dim_vec[:, j], *plot_args, **plot_kwargs)\n",
    "                        \n",
    "        for j in range(0, i):\n",
    "            axes[j, i].axis('off')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA1gAAANYCAYAAADZn0yoAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEwosAAAAAPDk/wPqrA54HptbagAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x11d0042b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "topic = 'documentary films'\n",
    "vec = vecs[interesting_subject_to_index[topic], None]\n",
    "\n",
    "fig, axes = plt.subplots(3, 3, figsize=(12, 12), sharex=True, sharey=True)\n",
    "\n",
    "plot_in_4d(axes, vecs, plot_kwargs={'color': 'lightgray'})\n",
    "plot_in_4d(axes, vec, plot_args=['xk'], label=topic)\n",
    "\n",
    "maybe_save_plot('2018-05-16-4d-plot')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I haven't sorted out why points like `documentary films` are so far away. I've seen suggestions that I should make the original data have unit variance. I didn't do it this time around because it seems trickier with the `sparse` matrices."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Highlighting a few points\n",
    "\n",
    "This is fun to explore! Here I plot a few different subjects. There seems to be one dimension for films, books, ending in a little corner for graphic novels and comic books. The other dimension stretches from children's books to other fiction to mystery and thriller books.\n",
    "\n",
    "Next, I can take one view and highlight a few different subjects."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABMIAAAR4CAYAAADkJnqzAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n+j0AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x117d78748>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fun_topics = [\n",
    "    'documentary films',\n",
    "    'fiction films',\n",
    "    'mystery fiction',\n",
    "    'thrillers fiction',\n",
    "    'audiobooks',\n",
    "    'friendship juvenile fiction',\n",
    "    'graphic novels',\n",
    "    'science fiction',\n",
    "    'large type books',\n",
    "    'television series',\n",
    "    'chinese language materials',\n",
    "    'picture books for children',\n",
    "    'serial murderers fiction',\n",
    "]\n",
    "\n",
    "fig, axes = plt.subplots(1, 1, figsize=(16, 16), sharex=True, sharey=True)\n",
    "\n",
    "x_dim = 2\n",
    "y_dim = 1\n",
    "\n",
    "axes.plot(vecs[:, x_dim], vecs[:, y_dim], '.', color='lightgray')\n",
    "\n",
    "for topic in fun_topics:\n",
    "    x = vecs[interesting_subject_to_index[topic], x_dim]\n",
    "    y = vecs[interesting_subject_to_index[topic], y_dim]    \n",
    "    axes.plot(x, y, 'xk')\n",
    "    axes.annotate(topic, (x - 0.002, y + 0.005), fontsize=14)\n",
    "\n",
    "maybe_save_plot('2018-05-16-labeled-plot')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Interactive\n",
    "\n",
    "I also tried making an interactive plot using `bokeh`. This is a little funky for now, but I'm posting it because it's really fun."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "from bokeh.models import HoverTool\n",
    "from bokeh.plotting import figure, show, output_file, ColumnDataSource\n",
    "from bokeh.layouts import gridplot\n",
    "from bokeh.palettes import magma\n",
    "\n",
    "def plot(u, x_dim='x1', y_dim='x2', descs=index_to_interesting_subject):\n",
    "\n",
    "    source = ColumnDataSource(data=dict(\n",
    "        x1=u[:, 0],\n",
    "        x2=u[:, 1],\n",
    "        x3=u[:, 2],\n",
    "        x4=u[:, 3],\n",
    "        desc=descs,\n",
    "    ))\n",
    "\n",
    "    hover = HoverTool(tooltips=[\n",
    "        (\"index\", \"$index\"),\n",
    "        (\"desc\", \"@desc\"),\n",
    "    ])\n",
    "\n",
    "    p = figure(plot_width=800, plot_height=800, tools=[hover, 'zoom_in', 'pan', 'zoom_out'])\n",
    "    p.scatter(\n",
    "        x_dim, y_dim,\n",
    "        size=5,\n",
    "        source=source,\n",
    "        line_color=None,\n",
    "        alpha=0.8,\n",
    "    )\n",
    "\n",
    "    grid = gridplot([[p]])\n",
    "\n",
    "    output_file(\"2018_05_16_library_catalog_subjects_{}_{}.html\".format(x_dim, y_dim), title=\"Library Catalog Subjects\")\n",
    "    \n",
    "    show(grid)\n",
    "\n",
    "plot(vecs, 'x1', 'x2')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## See Also\n",
    "\n",
    " - This data is also available on [data.seattle.gov](https://data.seattle.gov/Community/Checkouts-by-Title/tmmm-ytt6), which tells me that I should say:\n",
    "> \"The data made available here has been modified for use from its original source, which is the City of Seattle. Neither the City of Seattle nor the Office of the Chief Technology Officer (OCTO) makes any claims as to the completeness, timeliness, accuracy or content of any data contained in this application; makes any representation of any kind, including, but not limited to, warranty of the accuracy or fitness for a particular use; nor are any such warranties to be implied or inferred with respect to the information or data furnished herein. The data is subject to change as modifications and updates are complete. It is understood that the information contained in the web feed is being used at one's own risk.\"\n",
    "\n",
    " - I found it via [kaggle](https://www.kaggle.com/seattle-public-library/seattle-library-checkout-records)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}