http://git-wip-us.apache.org/repos/asf/madlib-site/blob/418f361c/community-artifacts/LDA-v1.ipynb
----------------------------------------------------------------------
diff --git a/community-artifacts/LDA-v1.ipynb b/community-artifacts/LDA-v1.ipynb
new file mode 100644
index 0000000..19a199c
--- /dev/null
+++ b/community-artifacts/LDA-v1.ipynb
@@ -0,0 +1,2034 @@
+{
+ "cells": [
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# Latent Dirichlet Allocation \n",
+ "\n",
+ "Latent Dirichlet Allocation (LDA) is a generative probabilistic model for
natural texts. It is used in problems such as automated topic discovery,
collaborative filtering, and document classification.\n",
+ "\n",
+ "In addition to an implementation of LDA, this MADlib module also provides
a number of additional helper functions to interpret results of the LDA output."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 1,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stderr",
+ "output_type": "stream",
+ "text": [
+
"/Users/fmcquillan/anaconda/lib/python2.7/site-packages/IPython/config.py:13:
ShimWarning: The `IPython.config` package has been deprecated. You should
import from traitlets.config instead.\n",
+ " \"You should import from traitlets.config instead.\", ShimWarning)\n",
+
"/Users/fmcquillan/anaconda/lib/python2.7/site-packages/IPython/utils/traitlets.py:5:
UserWarning: IPython.utils.traitlets has moved to a top-level traitlets
package.\n",
+ " warn(\"IPython.utils.traitlets has moved to a top-level traitlets
package.\")\n"
+ ]
+ }
+ ],
+ "source": [
+ "%load_ext sql"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 2,
+ "metadata": {},
+ "outputs": [
+ {
+ "data": {
+ "text/plain": [
+ "u'Connected: gpadmin@madlib'"
+ ]
+ },
+ "execution_count": 2,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "# Greenplum Database 5.4.0 on GCP (demo machine)\n",
+ "%sql postgresql://gpadmin@35.184.253.255:5432/madlib\n",
+ " \n",
+ "# PostgreSQL local\n",
+ "#%sql postgresql://fmcquillan@localhost:5432/madlib\n",
+ "\n",
+ "# Greenplum Database 4.3.10.0\n",
+ "#%sql postgresql://gpdbchina@10.194.10.68:61000/madlib"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 4,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "1 rows affected.\n"
+ ]
+ },
+ {
+ "data": {
+ "text/html": [
+ "<table>\n",
+ " <tr>\n",
+ " <th>version</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>MADlib version: 1.14-dev, git revision:
rc/1.13-rc1-15-g7ffad03, cmake configuration time: Wed Feb 21 01:33:31 UTC
2018, build type: release, build system: Linux-2.6.32-696.20.1.el6.x86_64, C
compiler: gcc 4.4.7, C++ compiler: g++ 4.4.7</td>\n",
+ " </tr>\n",
+ "</table>"
+ ],
+ "text/plain": [
+ "[(u'MADlib version: 1.14-dev, git revision: rc/1.13-rc1-15-g7ffad03,
cmake configuration time: Wed Feb 21 01:33:31 UTC 2018, build type: release,
build system: Linux-2.6.32-696.20.1.el6.x86_64, C compiler: gcc 4.4.7, C++
compiler: g++ 4.4.7',)]"
+ ]
+ },
+ "execution_count": 4,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "%sql select madlib.version();\n",
+ "#%sql select version();"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# 1. Prepare documents\n",
+ "The examples below are short strings extracted from various Wikipedia
documents. First we create a document table with one document per row:"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 5,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Done.\n",
+ "Done.\n",
+ "4 rows affected.\n",
+ "4 rows affected.\n"
+ ]
+ },
+ {
+ "data": {
+ "text/html": [
+ "<table>\n",
+ " <tr>\n",
+ " <th>docid</th>\n",
+ " <th>contents</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>Statistical topic models are a class of Bayesian latent
variable models, originally developed for analyzing the semantic content of
large document corpora.</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>By the late 1960s, the balance between pitching and
hitting had swung in favor of the pitchers. In 1968 Carl Yastrzemski won the
American League batting title with an average of just .301, the lowest in
history.</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>2</td>\n",
+ " <td>Machine learning is closely related to and often overlaps
with computational statistics; a discipline that also specializes in
prediction-making. It has strong ties to mathematical optimization, which
deliver methods, theory and application domains to the field.</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>3</td>\n",
+ " <td>California's diverse geography ranges from the Sierra
Nevada in the east to the Pacific Coast in the west, from the RedwoodâDouglas
fir forests of the northwest, to the Mojave Desert areas in the southeast. The
center of the state is dominated by the Central Valley, a major agricultural
area.</td>\n",
+ " </tr>\n",
+ "</table>"
+ ],
+ "text/plain": [
+ "[(0, u'Statistical topic models are a class of Bayesian latent
variable models, originally developed for analyzing the semantic content of
large document corpora.'),\n",
+ " (1, u'By the late 1960s, the balance between pitching and hitting had
swung in favor of the pitchers. In 1968 Carl Yastrzemski won the American
League batting title with an average of just .301, the lowest in history.'),\n",
+ " (2, u'Machine learning is closely related to and often overlaps with
computational statistics; a discipline that also specializes in
prediction-making. It has strong ties to mathematical optimization, which
deliver methods, theory and application domains to the field.'),\n",
+ " (3, u\"California's diverse geography ranges from the Sierra Nevada
in the east to the Pacific Coast in the west, from the Redwood\\u2013Douglas
fir forests of the northwest, to the Mojave Desert areas in the southeast. The
center of the state is dominated by the Central Valley, a major agricultural
area.\")]"
+ ]
+ },
+ "execution_count": 5,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "%%sql\n",
+ "DROP TABLE IF EXISTS documents;\n",
+ "CREATE TABLE documents(docid INT4, contents TEXT);\n",
+ "\n",
+ "INSERT INTO documents VALUES\n",
+ "(0, 'Statistical topic models are a class of Bayesian latent variable
models, originally developed for analyzing the semantic content of large
document corpora.'),\n",
+ "(1, 'By the late 1960s, the balance between pitching and hitting had
swung in favor of the pitchers. In 1968 Carl Yastrzemski won the American
League batting title with an average of just .301, the lowest in history.'),\n",
+ "(2, 'Machine learning is closely related to and often overlaps with
computational statistics; a discipline that also specializes in
prediction-making. It has strong ties to mathematical optimization, which
deliver methods, theory and application domains to the field.'),\n",
+ "(3, 'California''s diverse geography ranges from the Sierra Nevada in the
east to the Pacific Coast in the west, from the RedwoodâDouglas fir forests
of the northwest, to the Mojave Desert areas in the southeast. The center of
the state is dominated by the Central Valley, a major agricultural area.');\n",
+ "\n",
+ "SELECT * from documents ORDER BY docid;"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "You can apply stemming, stop word removal and tokenization at this point
in order to prepare the documents for text processing. Depending upon your
database version, various tools are available. Databases based on more recent
versions of PostgreSQL may do something like: "
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "metadata": {
+ "collapsed": true
+ },
+ "outputs": [],
+ "source": [
+ "%%sql\n",
+ "SELECT tsvector_to_array(to_tsvector('english',contents)) from documents;"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "In this example, we assume a database based on an older version of
PostgreSQL and just perform basic punctuation removal and tokenization. The
array of words is added as a new column to the documents table:"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 6,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Done.\n",
+ "4 rows affected.\n",
+ "4 rows affected.\n"
+ ]
+ },
+ {
+ "data": {
+ "text/html": [
+ "<table>\n",
+ " <tr>\n",
+ " <th>docid</th>\n",
+ " <th>contents</th>\n",
+ " <th>words</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>Statistical topic models are a class of Bayesian latent
variable models, originally developed for analyzing the semantic content of
large document corpora.</td>\n",
+ " <td>[u'statistical', u'topic', u'models', u'are', u'a',
u'class', u'of', u'bayesian', u'latent', u'variable', u'models', u'originally',
u'developed', u'for', u'analyzing', u'the', u'semantic', u'content', u'of',
u'large', u'document', u'corpora']</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>By the late 1960s, the balance between pitching and
hitting had swung in favor of the pitchers. In 1968 Carl Yastrzemski won the
American League batting title with an average of just .301, the lowest in
history.</td>\n",
+ " <td>[u'by', u'the', u'late', u'1960s', u'the', u'balance',
u'between', u'pitching', u'and', u'hitting', u'had', u'swung', u'in', u'favor',
u'of', u'the', u'pitchers', u'in', u'1968', u'carl', u'yastrzemski', u'won',
u'the', u'american', u'league', u'batting', u'title', u'with', u'an',
u'average', u'of', u'just', u'301', u'the', u'lowest', u'in',
u'history']</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>2</td>\n",
+ " <td>Machine learning is closely related to and often overlaps
with computational statistics; a discipline that also specializes in
prediction-making. It has strong ties to mathematical optimization, which
deliver methods, theory and application domains to the field.</td>\n",
+ " <td>[u'machine', u'learning', u'is', u'closely', u'related',
u'to', u'and', u'often', u'overlaps', u'with', u'computational', u'statistics',
u'a', u'discipline', u'that', u'also', u'specializes', u'in',
u'prediction-making', u'it', u'has', u'strong', u'ties', u'to',
u'mathematical', u'optimization', u'which', u'deliver', u'methods', u'theory',
u'and', u'application', u'domains', u'to', u'the', u'field']</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>3</td>\n",
+ " <td>California's diverse geography ranges from the Sierra
Nevada in the east to the Pacific Coast in the west, from the RedwoodâDouglas
fir forests of the northwest, to the Mojave Desert areas in the southeast. The
center of the state is dominated by the Central Valley, a major agricultural
area.</td>\n",
+ " <td>[u'californias', u'diverse', u'geography', u'ranges',
u'from', u'the', u'sierra', u'nevada', u'in', u'the', u'east', u'to', u'the',
u'pacific', u'coast', u'in', u'the', u'west', u'from', u'the',
u'redwood\\u2013douglas', u'fir', u'forests', u'of', u'the', u'northwest',
u'to', u'the', u'mojave', u'desert', u'areas', u'in', u'the', u'southeast',
u'the', u'center', u'of', u'the', u'state', u'is', u'dominated', u'by', u'the',
u'central', u'valley', u'a', u'major', u'agricultural', u'area']</td>\n",
+ " </tr>\n",
+ "</table>"
+ ],
+ "text/plain": [
+ "[(0, u'Statistical topic models are a class of Bayesian latent
variable models, originally developed for analyzing the semantic content of
large document corpora.', [u'statistical', u'topic', u'models', u'are', u'a',
u'class', u'of', u'bayesian', u'latent', u'variable', u'models', u'originally',
u'developed', u'for', u'analyzing', u'the', u'semantic', u'content', u'of',
u'large', u'document', u'corpora']),\n",
+ " (1, u'By the late 1960s, the balance between pitching and hitting had
swung in favor of the pitchers. In 1968 Carl Yastrzemski won the American
League batting title with an average of just .301, the lowest in history.',
[u'by', u'the', u'late', u'1960s', u'the', u'balance', u'between', u'pitching',
u'and', u'hitting', u'had', u'swung', u'in', u'favor', u'of', u'the',
u'pitchers', u'in', u'1968', u'carl', u'yastrzemski', u'won', u'the',
u'american', u'league', u'batting', u'title', u'with', u'an', u'average',
u'of', u'just', u'301', u'the', u'lowest', u'in', u'history']),\n",
+ " (2, u'Machine learning is closely related to and often overlaps with
computational statistics; a discipline that also specializes in
prediction-making. It has strong ties to mathematical optimization, which
deliver methods, theory and application domains to the field.', [u'machine',
u'learning', u'is', u'closely', u'related', u'to', u'and', u'often',
u'overlaps', u'with', u'computational', u'statistics', u'a', u'discipline',
u'that', u'also', u'specializes', u'in', u'prediction-making', u'it', u'has',
u'strong', u'ties', u'to', u'mathematical', u'optimization', u'which',
u'deliver', u'methods', u'theory', u'and', u'application', u'domains', u'to',
u'the', u'field']),\n",
+ " (3, u\"California's diverse geography ranges from the Sierra Nevada
in the east to the Pacific Coast in the west, from the Redwood\\u2013Douglas
fir forests of the northwest, to the Mojave Desert areas in the southeast. The
center of the state is dominated by the Central Valley, a major agricultural
area.\", [u'californias', u'diverse', u'geography', u'ranges', u'from', u'the',
u'sierra', u'nevada', u'in', u'the', u'east', u'to', u'the', u'pacific',
u'coast', u'in', u'the', u'west', u'from', u'the', u'redwood\\u2013douglas',
u'fir', u'forests', u'of', u'the', u'northwest', u'to', u'the', u'mojave',
u'desert', u'areas', u'in', u'the', u'southeast', u'the', u'center', u'of',
u'the', u'state', u'is', u'dominated', u'by', u'the', u'central', u'valley',
u'a', u'major', u'agricultural', u'area'])]"
+ ]
+ },
+ "execution_count": 6,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "%%sql\n",
+ "ALTER TABLE documents ADD COLUMN words TEXT[];\n",
+ "\n",
+ "UPDATE documents SET words = \n",
+ " regexp_split_to_array(lower(\n",
+ " regexp_replace(contents, E'[,.;\\']','', 'g')\n",
+ " ), E'[\\\\s+]');\n",
+ " \n",
+ "SELECT * FROM documents ORDER BY docid;"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# 2. Term frequency\n",
+ "Build a word count table by extracting the words and building a histogram
for each document using the term_frequency() function."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 7,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Done.\n",
+ "1 rows affected.\n",
+ "20 rows affected.\n"
+ ]
+ },
+ {
+ "data": {
+ "text/html": [
+ "<table>\n",
+ " <tr>\n",
+ " <th>docid</th>\n",
+ " <th>wordid</th>\n",
+ " <th>count</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>17</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>11</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>95</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>90</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>85</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>68</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>54</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>42</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>35</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>28</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>8</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>3</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>97</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>80</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>71</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>64</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>56</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>32</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>29</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>24</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ "</table>"
+ ],
+ "text/plain": [
+ "[(0, 17, 1),\n",
+ " (0, 11, 1),\n",
+ " (0, 95, 1),\n",
+ " (0, 90, 1),\n",
+ " (0, 85, 1),\n",
+ " (0, 68, 2),\n",
+ " (0, 54, 1),\n",
+ " (0, 42, 1),\n",
+ " (0, 35, 1),\n",
+ " (0, 28, 1),\n",
+ " (0, 8, 1),\n",
+ " (0, 3, 1),\n",
+ " (0, 97, 1),\n",
+ " (0, 80, 1),\n",
+ " (0, 71, 1),\n",
+ " (0, 64, 2),\n",
+ " (0, 56, 1),\n",
+ " (0, 32, 1),\n",
+ " (0, 29, 1),\n",
+ " (0, 24, 1)]"
+ ]
+ },
+ "execution_count": 7,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "%%sql\n",
+ "DROP TABLE IF EXISTS documents_tf, documents_tf_vocabulary;\n",
+ "\n",
+ "SELECT madlib.term_frequency('documents', -- input table\n",
+ " 'docid', -- document id column\n",
+ " 'words', -- vector of words in
document\n",
+ " 'documents_tf', -- output documents table
with term frequency\n",
+ " TRUE); -- TRUE to created
vocabulary table\n",
+ "\n",
+ "SELECT * FROM documents_tf ORDER BY docid LIMIT 20;"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "Here is the associated vocabulary table. Note that wordid starts at 0:"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 8,
+ "metadata": {
+ "scrolled": true
+ },
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "20 rows affected.\n"
+ ]
+ },
+ {
+ "data": {
+ "text/html": [
+ "<table>\n",
+ " <tr>\n",
+ " <th>wordid</th>\n",
+ " <th>word</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>1960s</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>1968</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>2</td>\n",
+ " <td>301</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>3</td>\n",
+ " <td>a</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>4</td>\n",
+ " <td>agricultural</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>5</td>\n",
+ " <td>also</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>6</td>\n",
+ " <td>american</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>7</td>\n",
+ " <td>an</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>8</td>\n",
+ " <td>analyzing</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>9</td>\n",
+ " <td>and</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>10</td>\n",
+ " <td>application</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>11</td>\n",
+ " <td>are</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>12</td>\n",
+ " <td>area</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>13</td>\n",
+ " <td>areas</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>14</td>\n",
+ " <td>average</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>15</td>\n",
+ " <td>balance</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>16</td>\n",
+ " <td>batting</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>17</td>\n",
+ " <td>bayesian</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>18</td>\n",
+ " <td>between</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>19</td>\n",
+ " <td>by</td>\n",
+ " </tr>\n",
+ "</table>"
+ ],
+ "text/plain": [
+ "[(0, u'1960s'),\n",
+ " (1, u'1968'),\n",
+ " (2, u'301'),\n",
+ " (3, u'a'),\n",
+ " (4, u'agricultural'),\n",
+ " (5, u'also'),\n",
+ " (6, u'american'),\n",
+ " (7, u'an'),\n",
+ " (8, u'analyzing'),\n",
+ " (9, u'and'),\n",
+ " (10, u'application'),\n",
+ " (11, u'are'),\n",
+ " (12, u'area'),\n",
+ " (13, u'areas'),\n",
+ " (14, u'average'),\n",
+ " (15, u'balance'),\n",
+ " (16, u'batting'),\n",
+ " (17, u'bayesian'),\n",
+ " (18, u'between'),\n",
+ " (19, u'by')]"
+ ]
+ },
+ "execution_count": 8,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "%%sql\n",
+ "SELECT * FROM documents_tf_vocabulary ORDER BY wordid LIMIT 20;"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "The total number of words in the vocabulary across all documents is:"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 9,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "1 rows affected.\n"
+ ]
+ },
+ {
+ "data": {
+ "text/html": [
+ "<table>\n",
+ " <tr>\n",
+ " <th>count</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>103</td>\n",
+ " </tr>\n",
+ "</table>"
+ ],
+ "text/plain": [
+ "[(103L,)]"
+ ]
+ },
+ "execution_count": 9,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "%%sql\n",
+ "SELECT COUNT(*) FROM documents_tf_vocabulary;"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# 3. Train LDA model\n",
+ "For Dirichlet priors we use initial rule-of-thumb values of 50/(number of
topics) for alpha and 0.01 for beta.\n",
+ "\n",
+ "Reminder that column names for docid, wordid, and count are currently
fixed, so you must use these exact names in the input table. After a successful
run of the LDA training function two tables are generated, one for storing the
learned model and the other for storing the output data table."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 10,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Done.\n",
+ "2 rows affected.\n",
+ "4 rows affected.\n"
+ ]
+ },
+ {
+ "data": {
+ "text/html": [
+ "<table>\n",
+ " <tr>\n",
+ " <th>docid</th>\n",
+ " <th>wordcount</th>\n",
+ " <th>words</th>\n",
+ " <th>counts</th>\n",
+ " <th>topic_count</th>\n",
+ " <th>topic_assignment</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>22</td>\n",
+ " <td>[24, 17, 11, 95, 90, 85, 68, 54, 42, 35, 28, 8, 3, 97, 80,
71, 64, 56, 32, 29]</td>\n",
+ " <td>[1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1,
1]</td>\n",
+ " <td>[3, 6, 5, 7, 1]</td>\n",
+ " <td>[3, 3, 1, 3, 0, 0, 2, 1, 1, 2, 1, 4, 2, 0, 1, 2, 3, 3, 3,
1, 2, 3]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>37</td>\n",
+ " <td>[1, 50, 49, 46, 19, 16, 14, 9, 7, 0, 90, 68, 57, 102, 101,
100, 93, 88, 75, 74, 59, 55, 53, 48, 39, 21, 18, 15, 6, 2]</td>\n",
+ " <td>[1, 3, 1, 1, 1, 1, 1, 1, 1, 1, 5, 2, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]</td>\n",
+ " <td>[12, 8, 3, 10, 4]</td>\n",
+ " <td>[0, 0, 0, 0, 1, 3, 0, 1, 1, 1, 3, 1, 0, 0, 0, 3, 3, 2, 1,
4, 4, 2, 2, 3, 1, 1, 3, 3, 3, 3, 0, 4, 3, 0, 4, 0, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>2</td>\n",
+ " <td>36</td>\n",
+ " <td>[10, 27, 33, 40, 47, 51, 58, 62, 63, 69, 72, 83, 100, 99,
94, 92, 91, 90, 89, 87, 86, 79, 76, 70, 60, 52, 50, 36, 30, 25, 9, 5,
3]</td>\n",
+ " <td>[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 3, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1]</td>\n",
+ " <td>[9, 12, 5, 3, 7]</td>\n",
+ " <td>[0, 0, 1, 4, 4, 4, 3, 4, 2, 2, 3, 2, 0, 1, 0, 0, 0, 2, 1,
3, 1, 1, 1, 4, 4, 2, 1, 4, 0, 1, 1, 1, 1, 1, 0, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>3</td>\n",
+ " <td>49</td>\n",
+ " <td>[77, 78, 81, 82, 67, 65, 51, 45, 44, 43, 34, 26, 13, 98,
96, 94, 90, 84, 73, 68, 66, 61, 50, 41, 38, 37, 31, 23, 22, 20, 19, 12, 4,
3]</td>\n",
+ " <td>[1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 2, 11, 1, 1,
2, 1, 1, 3, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]</td>\n",
+ " <td>[16, 5, 7, 13, 8]</td>\n",
+ " <td>[2, 1, 3, 4, 1, 4, 0, 0, 4, 4, 3, 2, 1, 4, 3, 4, 0, 0, 3,
3, 0, 0, 3, 3, 0, 3, 0, 3, 0, 4, 3, 1, 1, 3, 3, 0, 0, 0, 2, 0, 2, 2, 0, 2, 3,
0, 2, 4, 0]</td>\n",
+ " </tr>\n",
+ "</table>"
+ ],
+ "text/plain": [
+ "[(0, 22, [24, 17, 11, 95, 90, 85, 68, 54, 42, 35, 28, 8, 3, 97, 80,
71, 64, 56, 32, 29], [1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1,
1], [3, 6, 5, 7, 1], [3, 3, 1, 3, 0, 0, 2, 1, 1, 2, 1, 4, 2, 0, 1, 2, 3, 3, 3,
1, 2, 3]),\n",
+ " (1, 37, [1, 50, 49, 46, 19, 16, 14, 9, 7, 0, 90, 68, 57, 102, 101,
100, 93, 88, 75, 74, 59, 55, 53, 48, 39, 21, 18, 15, 6, 2], [1, 3, 1, 1, 1, 1,
1, 1, 1, 1, 5, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [12,
8, 3, 10, 4], [0, 0, 0, 0, 1, 3, 0, 1, 1, 1, 3, 1, 0, 0, 0, 3, 3, 2, 1, 4, 4,
2, 2, 3, 1, 1, 3, 3, 3, 3, 0, 4, 3, 0, 4, 0, 0]),\n",
+ " (2, 36, [10, 27, 33, 40, 47, 51, 58, 62, 63, 69, 72, 83, 100, 99, 94,
92, 91, 90, 89, 87, 86, 79, 76, 70, 60, 52, 50, 36, 30, 25, 9, 5, 3], [1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 3, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 2, 1, 1], [9, 12, 5, 3, 7], [0, 0, 1, 4, 4, 4, 3, 4, 2, 2, 3, 2, 0, 1, 0, 0,
0, 2, 1, 3, 1, 1, 1, 4, 4, 2, 1, 4, 0, 1, 1, 1, 1, 1, 0, 0]),\n",
+ " (3, 49, [77, 78, 81, 82, 67, 65, 51, 45, 44, 43, 34, 26, 13, 98, 96,
94, 90, 84, 73, 68, 66, 61, 50, 41, 38, 37, 31, 23, 22, 20, 19, 12, 4, 3], [1,
1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 2, 11, 1, 1, 2, 1, 1, 3, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1], [16, 5, 7, 13, 8], [2, 1, 3, 4, 1, 4, 0, 0, 4, 4, 3, 2,
1, 4, 3, 4, 0, 0, 3, 3, 0, 0, 3, 3, 0, 3, 0, 3, 0, 4, 3, 1, 1, 3, 3, 0, 0, 0,
2, 0, 2, 2, 0, 2, 3, 0, 2, 4, 0])]"
+ ]
+ },
+ "execution_count": 10,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "%%sql\n",
+ "DROP TABLE IF EXISTS lda_model, lda_output_data;\n",
+ "\n",
+ "SELECT madlib.lda_train( 'documents_tf', -- documents table in the
form of term frequency\n",
+ " 'lda_model', -- model table created by
LDA training (not human readable)\n",
+ " 'lda_output_data', -- readable output data
table \n",
+ " 103, -- vocabulary size\n",
+ " 5, -- number of topics\n",
+ " 10, -- number of iterations\n",
+ " 5, -- Dirichlet prior for the
per-doc topic multinomial (alpha)\n",
+ " 0.01 -- Dirichlet prior for the
per-topic word multinomial (beta)\n",
+ " );\n",
+ "\n",
+ "SELECT * FROM lda_output_data ORDER BY docid;"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# 4. Helper functions on learned model \n",
+ "\n",
+ "First, we get topic description by top-k words. These are the k words
with the highest probability for the topic.\n",
+ "\n",
+ "Note that if there are ties in probability, more than k words may
actually be reported for each topic. Also note that topicid starts at 0."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 11,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Done.\n",
+ "1 rows affected.\n",
+ "40 rows affected.\n"
+ ]
+ },
+ {
+ "data": {
+ "text/html": [
+ "<table>\n",
+ " <tr>\n",
+ " <th>topicid</th>\n",
+ " <th>wordid</th>\n",
+ " <th>prob</th>\n",
+ " <th>word</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>90</td>\n",
+ " <td>0.219595417987</td>\n",
+ " <td>the</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>50</td>\n",
+ " <td>0.170850597124</td>\n",
+ " <td>in</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>94</td>\n",
+ " <td>0.122105776261</td>\n",
+ " <td>to</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>3</td>\n",
+ " <td>0.0733609553985</td>\n",
+ " <td>a</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>19</td>\n",
+ " <td>0.0489885449671</td>\n",
+ " <td>by</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>68</td>\n",
+ " <td>0.125195129566</td>\n",
+ " <td>of</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>9</td>\n",
+ " <td>0.0939743990009</td>\n",
+ " <td>and</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>36</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>domains</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>14</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>average</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>16</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>batting</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>54</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>large</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>56</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>latent</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>78</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>redwoodâdouglas</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>86</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>statistics</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>60</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>machine</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>26</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>coast</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>0</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>1960s</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>25</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>closely</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>87</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>strong</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>67</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>northwest</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>99</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>which</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>35</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>document</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>11</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>are</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>91</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>theory</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>33</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>discipline</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>75</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>pitching</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>49</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>hitting</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>97</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>variable</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>89</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>that</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>88</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>swung</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>30</td>\n",
+ " <td>0.0315329378707</td>\n",
+ " <td>deliver</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>2</td>\n",
+ " <td>68</td>\n",
+ " <td>0.095577746077</td>\n",
+ " <td>of</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>2</td>\n",
+ " <td>101</td>\n",
+ " <td>0.0480266286258</td>\n",
+ " <td>won</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>2</td>\n",
+ " <td>63</td>\n",
+ " <td>0.0480266286258</td>\n",
+ " <td>methods</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>2</td>\n",
+ " <td>41</td>\n",
+ " <td>0.0480266286258</td>\n",
+ " <td>fir</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>2</td>\n",
+ " <td>83</td>\n",
+ " <td>0.0480266286258</td>\n",
+ " <td>specializes</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>2</td>\n",
+ " <td>77</td>\n",
+ " <td>0.0480266286258</td>\n",
+ " <td>ranges</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>2</td>\n",
+ " <td>37</td>\n",
+ " <td>0.0480266286258</td>\n",
+ " <td>dominated</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>2</td>\n",
+ " <td>100</td>\n",
+ " <td>0.0480266286258</td>\n",
+ " <td>with</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>2</td>\n",
+ " <td>22</td>\n",
+ " <td>0.0480266286258</td>\n",
+ " <td>center</td>\n",
+ " </tr>\n",
+ "</table>"
+ ],
+ "text/plain": [
+ "[(0, 90, 0.219595417986839, u'the'),\n",
+ " (0, 50, 0.170850597124056, u'in'),\n",
+ " (0, 94, 0.122105776261272, u'to'),\n",
+ " (0, 3, 0.0733609553984889, u'a'),\n",
+ " (0, 19, 0.0489885449670972, u'by'),\n",
+ " (1, 68, 0.125195129566032, u'of'),\n",
+ " (1, 9, 0.0939743990009366, u'and'),\n",
+ " (1, 36, 0.0315329378707462, u'domains'),\n",
+ " (1, 14, 0.0315329378707462, u'average'),\n",
+ " (1, 16, 0.0315329378707462, u'batting'),\n",
+ " (1, 54, 0.0315329378707462, u'large'),\n",
+ " (1, 56, 0.0315329378707462, u'latent'),\n",
+ " (1, 78, 0.0315329378707462, u'redwood\\u2013douglas'),\n",
+ " (1, 86, 0.0315329378707462, u'statistics'),\n",
+ " (1, 60, 0.0315329378707462, u'machine'),\n",
+ " (1, 26, 0.0315329378707462, u'coast'),\n",
+ " (1, 0, 0.0315329378707462, u'1960s'),\n",
+ " (1, 25, 0.0315329378707462, u'closely'),\n",
+ " (1, 87, 0.0315329378707462, u'strong'),\n",
+ " (1, 67, 0.0315329378707462, u'northwest'),\n",
+ " (1, 99, 0.0315329378707462, u'which'),\n",
+ " (1, 35, 0.0315329378707462, u'document'),\n",
+ " (1, 11, 0.0315329378707462, u'are'),\n",
+ " (1, 91, 0.0315329378707462, u'theory'),\n",
+ " (1, 33, 0.0315329378707462, u'discipline'),\n",
+ " (1, 75, 0.0315329378707462, u'pitching'),\n",
+ " (1, 49, 0.0315329378707462, u'hitting'),\n",
+ " (1, 97, 0.0315329378707462, u'variable'),\n",
+ " (1, 89, 0.0315329378707462, u'that'),\n",
+ " (1, 88, 0.0315329378707462, u'swung'),\n",
+ " (1, 30, 0.0315329378707462, u'deliver'),\n",
+ " (2, 68, 0.0955777460770328, u'of'),\n",
+ " (2, 101, 0.0480266286257727, u'won'),\n",
+ " (2, 63, 0.0480266286257727, u'methods'),\n",
+ " (2, 41, 0.0480266286257727, u'fir'),\n",
+ " (2, 83, 0.0480266286257727, u'specializes'),\n",
+ " (2, 77, 0.0480266286257727, u'ranges'),\n",
+ " (2, 37, 0.0480266286257727, u'dominated'),\n",
+ " (2, 100, 0.0480266286257727, u'with'),\n",
+ " (2, 22, 0.0480266286257727, u'center')]"
+ ]
+ },
+ "execution_count": 11,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "%%sql\n",
+ "DROP TABLE IF EXISTS helper_output_table;\n",
+ "\n",
+ "SELECT madlib.lda_get_topic_desc( 'lda_model', -- LDA
model generated in training\n",
+ " 'documents_tf_vocabulary', --
vocabulary table that maps wordid to word\n",
+ " 'helper_output_table', -- output
table for per-topic descriptions\n",
+ " 5); -- k:
number of top words for each topic\n",
+ "\n",
+ "SELECT * FROM helper_output_table ORDER BY topicid, prob DESC LIMIT 40;"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "Get the per-word topic counts. This mapping shows how many times a given
word is assigned to a topic. E.g., wordid 3 is assigned to topicid 0 three
times. "
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 12,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Done.\n",
+ "1 rows affected.\n",
+ "20 rows affected.\n"
+ ]
+ },
+ {
+ "data": {
+ "text/html": [
+ "<table>\n",
+ " <tr>\n",
+ " <th>wordid</th>\n",
+ " <th>topic_count</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>[0, 1, 0, 0, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>[1, 0, 0, 0, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>2</td>\n",
+ " <td>[1, 0, 0, 0, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>3</td>\n",
+ " <td>[3, 0, 0, 0, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>4</td>\n",
+ " <td>[0, 0, 0, 0, 1]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>5</td>\n",
+ " <td>[1, 0, 0, 0, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>6</td>\n",
+ " <td>[1, 0, 0, 0, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>7</td>\n",
+ " <td>[0, 0, 0, 1, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>8</td>\n",
+ " <td>[0, 0, 1, 0, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>9</td>\n",
+ " <td>[0, 3, 0, 0, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>10</td>\n",
+ " <td>[1, 0, 0, 0, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>11</td>\n",
+ " <td>[0, 1, 0, 0, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>12</td>\n",
+ " <td>[0, 0, 1, 0, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>13</td>\n",
+ " <td>[0, 0, 0, 0, 1]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>14</td>\n",
+ " <td>[0, 1, 0, 0, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>15</td>\n",
+ " <td>[0, 0, 0, 0, 1]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>16</td>\n",
+ " <td>[0, 1, 0, 0, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>17</td>\n",
+ " <td>[0, 0, 0, 1, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>18</td>\n",
+ " <td>[1, 0, 0, 0, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>19</td>\n",
+ " <td>[2, 0, 0, 0, 0]</td>\n",
+ " </tr>\n",
+ "</table>"
+ ],
+ "text/plain": [
+ "[(0, [0, 1, 0, 0, 0]),\n",
+ " (1, [1, 0, 0, 0, 0]),\n",
+ " (2, [1, 0, 0, 0, 0]),\n",
+ " (3, [3, 0, 0, 0, 0]),\n",
+ " (4, [0, 0, 0, 0, 1]),\n",
+ " (5, [1, 0, 0, 0, 0]),\n",
+ " (6, [1, 0, 0, 0, 0]),\n",
+ " (7, [0, 0, 0, 1, 0]),\n",
+ " (8, [0, 0, 1, 0, 0]),\n",
+ " (9, [0, 3, 0, 0, 0]),\n",
+ " (10, [1, 0, 0, 0, 0]),\n",
+ " (11, [0, 1, 0, 0, 0]),\n",
+ " (12, [0, 0, 1, 0, 0]),\n",
+ " (13, [0, 0, 0, 0, 1]),\n",
+ " (14, [0, 1, 0, 0, 0]),\n",
+ " (15, [0, 0, 0, 0, 1]),\n",
+ " (16, [0, 1, 0, 0, 0]),\n",
+ " (17, [0, 0, 0, 1, 0]),\n",
+ " (18, [1, 0, 0, 0, 0]),\n",
+ " (19, [2, 0, 0, 0, 0])]"
+ ]
+ },
+ "execution_count": 12,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "%%sql\n",
+ "DROP TABLE IF EXISTS helper_output_table;\n",
+ "\n",
+ "SELECT madlib.lda_get_word_topic_count( 'lda_model', -- LDA
model generated in training\n",
+ " 'helper_output_table'); -- output
table for per-word topic counts\n",
+ "\n",
+ "SELECT * FROM helper_output_table ORDER BY wordid LIMIT 20;"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "Get the per-topic word counts. This mapping shows which words are
associated with each topic by frequency."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 13,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Done.\n",
+ "1 rows affected.\n",
+ "5 rows affected.\n"
+ ]
+ },
+ {
+ "data": {
+ "text/html": [
+ "<table>\n",
+ " <tr>\n",
+ " <th>topicid</th>\n",
+ " <th>word_count</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>[0, 1, 1, 3, 0, 1, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1,
2, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0,
1, 0, 0, 1, 0, 7, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 9, 0, 0, 0, 5, 0, 0,
0, 0, 0, 1, 0, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>[1, 0, 0, 0, 0, 0, 0, 0, 0, 3, 0, 1, 0, 0, 1, 0, 1, 0, 0,
0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 1, 0, 0, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 4, 0, 0,
0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 1, 0, 0, 0, 0, 0,
1, 0, 1, 0, 0, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>2</td>\n",
+ " <td>[0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0,
0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 0, 1, 0, 0, 0, 1, 1, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 2, 1, 1,
0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0,
0, 0, 0, 1, 1, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>3</td>\n",
+ " <td>[0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0,
0, 1, 1, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0,
0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 1, 1, 0, 1, 0, 0, 2, 0, 1, 0, 0, 0, 0,
1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 9, 0, 0, 1, 0, 1, 0,
0, 1, 0, 0, 0, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>4</td>\n",
+ " <td>[0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 2,
0, 0, 1, 0, 0, 0, 1, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1,
0, 0, 0, 0, 0, 1]</td>\n",
+ " </tr>\n",
+ "</table>"
+ ],
+ "text/plain": [
+ "[(0, [0, 1, 1, 3, 0, 1, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 2, 0,
0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0,
0, 1, 0, 7, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 9, 0, 0, 0, 5, 0, 0, 0, 0,
0, 1, 0, 0]),\n",
+ " (1, [1, 0, 0, 0, 0, 0, 0, 0, 0, 3, 0, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0,
0, 0, 0, 0, 1, 1, 0, 0, 0, 1, 0, 0, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 1, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 4, 0, 0, 0, 0,
0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 1, 0, 0, 0, 0, 0, 1, 0,
1, 0, 0, 0]),\n",
+ " (2, [0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0,
0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 0, 1, 0, 0, 0, 1, 1, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 2, 1, 1, 0, 0,
0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0,
0, 1, 1, 0]),\n",
+ " (3, [0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1,
1, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1,
0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 1, 1, 0, 1, 0, 0, 2, 0, 1, 0, 0, 0, 0, 1, 1,
1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 9, 0, 0, 1, 0, 1, 0, 0, 1,
0, 0, 0, 0]),\n",
+ " (4, [0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 2, 0, 0,
1, 0, 0, 0, 1, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 1, 0, 0, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0,
0, 0, 0, 1])]"
+ ]
+ },
+ "execution_count": 13,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "%%sql\n",
+ "DROP TABLE IF EXISTS helper_output_table;\n",
+ "\n",
+ "SELECT madlib.lda_get_topic_word_count( 'lda_model',\n",
+ " 'helper_output_table');\n",
+ "\n",
+ "SELECT * FROM helper_output_table ORDER BY topicid;"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "Get the per-document word to topic mapping:"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 14,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Done.\n",
+ "1 rows affected.\n",
+ "40 rows affected.\n"
+ ]
+ },
+ {
+ "data": {
+ "text/html": [
+ "<table>\n",
+ " <tr>\n",
+ " <th>docid</th>\n",
+ " <th>wordid</th>\n",
+ " <th>topicid</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>54</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>42</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>35</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>32</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>29</td>\n",
+ " <td>3</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>28</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>24</td>\n",
+ " <td>3</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>17</td>\n",
+ " <td>3</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>11</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>8</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>3</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>97</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>95</td>\n",
+ " <td>3</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>90</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>85</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>80</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>71</td>\n",
+ " <td>3</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>68</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>68</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>64</td>\n",
+ " <td>3</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>56</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>1</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>0</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>102</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>101</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>100</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>93</td>\n",
+ " <td>3</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>90</td>\n",
+ " <td>3</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>90</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>88</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>75</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>74</td>\n",
+ " <td>3</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>68</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>68</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>59</td>\n",
+ " <td>3</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>57</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>55</td>\n",
+ " <td>3</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>53</td>\n",
+ " <td>3</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>50</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>49</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ "</table>"
+ ],
+ "text/plain": [
+ "[(0, 54, 1),\n",
+ " (0, 42, 2),\n",
+ " (0, 35, 1),\n",
+ " (0, 32, 2),\n",
+ " (0, 29, 3),\n",
+ " (0, 28, 4),\n",
+ " (0, 24, 3),\n",
+ " (0, 17, 3),\n",
+ " (0, 11, 1),\n",
+ " (0, 8, 2),\n",
+ " (0, 3, 0),\n",
+ " (0, 97, 1),\n",
+ " (0, 95, 3),\n",
+ " (0, 90, 0),\n",
+ " (0, 85, 0),\n",
+ " (0, 80, 2),\n",
+ " (0, 71, 3),\n",
+ " (0, 68, 2),\n",
+ " (0, 68, 1),\n",
+ " (0, 64, 3),\n",
+ " (0, 56, 1),\n",
+ " (1, 1, 0),\n",
+ " (1, 0, 1),\n",
+ " (1, 102, 4),\n",
+ " (1, 101, 2),\n",
+ " (1, 100, 2),\n",
+ " (1, 93, 3),\n",
+ " (1, 90, 3),\n",
+ " (1, 90, 0),\n",
+ " (1, 88, 1),\n",
+ " (1, 75, 1),\n",
+ " (1, 74, 3),\n",
+ " (1, 68, 2),\n",
+ " (1, 68, 1),\n",
+ " (1, 59, 3),\n",
+ " (1, 57, 4),\n",
+ " (1, 55, 3),\n",
+ " (1, 53, 3),\n",
+ " (1, 50, 0),\n",
+ " (1, 49, 1)]"
+ ]
+ },
+ "execution_count": 14,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "%%sql\n",
+ "DROP TABLE IF EXISTS helper_output_table;\n",
+ "\n",
+ "SELECT madlib.lda_get_word_topic_mapping('lda_output_data',\n",
+ " 'helper_output_table');\n",
+ "\n",
+ "SELECT * FROM helper_output_table ORDER BY docid LIMIT 40;"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# 6. Predict\n",
+ "Use a learned LDA model for prediction (that is, to label new documents).
In this example, we use the same input table as we used to train, just for
demonstration purpose. Normally, the test document is a new one that we want
to predict on."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 15,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Done.\n",
+ "1 rows affected.\n",
+ "4 rows affected.\n"
+ ]
+ },
+ {
+ "data": {
+ "text/html": [
+ "<table>\n",
+ " <tr>\n",
+ " <th>docid</th>\n",
+ " <th>wordcount</th>\n",
+ " <th>words</th>\n",
+ " <th>counts</th>\n",
+ " <th>topic_count</th>\n",
+ " <th>topic_assignment</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>37</td>\n",
+ " <td>[1, 50, 49, 46, 19, 16, 14, 9, 7, 0, 90, 68, 57, 102, 101,
100, 93, 88, 75, 74, 59, 55, 53, 48, 39, 21, 18, 15, 6, 2]</td>\n",
+ " <td>[1, 3, 1, 1, 1, 1, 1, 1, 1, 1, 5, 2, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]</td>\n",
+ " <td>[7, 3, 19, 2, 6]</td>\n",
+ " <td>[2, 0, 0, 0, 2, 2, 2, 2, 2, 1, 2, 2, 3, 3, 0, 0, 0, 1, 1,
2, 2, 0, 4, 2, 2, 2, 4, 4, 4, 4, 2, 2, 2, 4, 2, 2, 2]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>3</td>\n",
+ " <td>49</td>\n",
+ " <td>[77, 78, 81, 82, 67, 65, 51, 45, 44, 43, 34, 26, 13, 98,
96, 94, 90, 84, 73, 68, 66, 61, 50, 41, 38, 37, 31, 23, 22, 20, 19, 12, 4,
3]</td>\n",
+ " <td>[1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 2, 11, 1, 1,
2, 1, 1, 3, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]</td>\n",
+ " <td>[20, 0, 3, 5, 21]</td>\n",
+ " <td>[4, 4, 4, 4, 0, 0, 3, 4, 4, 3, 4, 4, 0, 4, 4, 2, 0, 0, 0,
0, 0, 3, 0, 0, 0, 0, 3, 3, 0, 0, 4, 2, 2, 4, 4, 0, 0, 0, 4, 4, 4, 4, 4, 4, 4,
0, 0, 4, 0]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>22</td>\n",
+ " <td>[24, 17, 11, 95, 90, 85, 68, 54, 42, 35, 28, 8, 3, 97, 80,
71, 64, 56, 32, 29]</td>\n",
+ " <td>[1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1,
1]</td>\n",
+ " <td>[0, 1, 3, 3, 15]</td>\n",
+ " <td>[4, 4, 4, 2, 3, 3, 2, 1, 4, 4, 4, 2, 4, 4, 4, 4, 4, 3, 4,
4, 4, 4]</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>2</td>\n",
+ " <td>36</td>\n",
+ " <td>[10, 27, 33, 40, 47, 51, 58, 62, 63, 69, 72, 83, 100, 99,
94, 92, 91, 90, 89, 87, 86, 79, 76, 70, 60, 52, 50, 36, 30, 25, 9, 5,
3]</td>\n",
+ " <td>[1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 3, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1]</td>\n",
+ " <td>[7, 0, 22, 1, 6]</td>\n",
+ " <td>[2, 2, 2, 2, 2, 2, 4, 2, 0, 2, 2, 2, 2, 2, 0, 0, 0, 2, 0,
3, 4, 2, 4, 2, 2, 2, 4, 2, 0, 4, 2, 4, 2, 2, 2, 0]</td>\n",
+ " </tr>\n",
+ "</table>"
+ ],
+ "text/plain": [
+ "[(1, 37, [1, 50, 49, 46, 19, 16, 14, 9, 7, 0, 90, 68, 57, 102, 101,
100, 93, 88, 75, 74, 59, 55, 53, 48, 39, 21, 18, 15, 6, 2], [1, 3, 1, 1, 1, 1,
1, 1, 1, 1, 5, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [7, 3,
19, 2, 6], [2, 0, 0, 0, 2, 2, 2, 2, 2, 1, 2, 2, 3, 3, 0, 0, 0, 1, 1, 2, 2, 0,
4, 2, 2, 2, 4, 4, 4, 4, 2, 2, 2, 4, 2, 2, 2]),\n",
+ " (3, 49, [77, 78, 81, 82, 67, 65, 51, 45, 44, 43, 34, 26, 13, 98, 96,
94, 90, 84, 73, 68, 66, 61, 50, 41, 38, 37, 31, 23, 22, 20, 19, 12, 4, 3], [1,
1, 1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 2, 11, 1, 1, 2, 1, 1, 3, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1], [20, 0, 3, 5, 21], [4, 4, 4, 4, 0, 0, 3, 4, 4, 3, 4, 4,
0, 4, 4, 2, 0, 0, 0, 0, 0, 3, 0, 0, 0, 0, 3, 3, 0, 0, 4, 2, 2, 4, 4, 0, 0, 0,
4, 4, 4, 4, 4, 4, 4, 0, 0, 4, 0]),\n",
+ " (0, 22, [24, 17, 11, 95, 90, 85, 68, 54, 42, 35, 28, 8, 3, 97, 80,
71, 64, 56, 32, 29], [1, 1, 1, 1, 1, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 1, 1,
1], [0, 1, 3, 3, 15], [4, 4, 4, 2, 3, 3, 2, 1, 4, 4, 4, 2, 4, 4, 4, 4, 4, 3, 4,
4, 4, 4]),\n",
+ " (2, 36, [10, 27, 33, 40, 47, 51, 58, 62, 63, 69, 72, 83, 100, 99, 94,
92, 91, 90, 89, 87, 86, 79, 76, 70, 60, 52, 50, 36, 30, 25, 9, 5, 3], [1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 3, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 2, 1, 1], [7, 0, 22, 1, 6], [2, 2, 2, 2, 2, 2, 4, 2, 0, 2, 2, 2, 2, 2, 0, 0,
0, 2, 0, 3, 4, 2, 4, 2, 2, 2, 4, 2, 0, 4, 2, 4, 2, 2, 2, 0])]"
+ ]
+ },
+ "execution_count": 15,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "%%sql\n",
+ "DROP TABLE IF EXISTS outdata_predict;\n",
+ "\n",
+ "SELECT madlib.lda_predict( 'documents_tf', -- Document to
predict\n",
+ " 'lda_model', -- LDA model from
training\n",
+ " 'outdata_predict' \n",
+ " );\n",
+ "\n",
+ "SELECT * FROM outdata_predict;"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# 7. Helper function on prediction"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 16,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Done.\n",
+ "1 rows affected.\n",
+ "40 rows affected.\n"
+ ]
+ },
+ {
+ "data": {
+ "text/html": [
+ "<table>\n",
+ " <tr>\n",
+ " <th>docid</th>\n",
+ " <th>wordid</th>\n",
+ " <th>topicid</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>42</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>35</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>32</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>29</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>28</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>24</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>17</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>11</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>8</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>3</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>97</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>95</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>90</td>\n",
+ " <td>3</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>85</td>\n",
+ " <td>3</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>80</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>71</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>68</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>68</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>64</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>64</td>\n",
+ " <td>3</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>56</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>0</td>\n",
+ " <td>54</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>6</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>2</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>1</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>0</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>102</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>101</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>100</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>93</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>90</td>\n",
+ " <td>3</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>90</td>\n",
+ " <td>0</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>88</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>75</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>74</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>68</td>\n",
+ " <td>1</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>59</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>57</td>\n",
+ " <td>2</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>55</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>1</td>\n",
+ " <td>53</td>\n",
+ " <td>4</td>\n",
+ " </tr>\n",
+ "</table>"
+ ],
+ "text/plain": [
+ "[(0, 42, 4),\n",
+ " (0, 35, 4),\n",
+ " (0, 32, 4),\n",
+ " (0, 29, 4),\n",
+ " (0, 28, 2),\n",
+ " (0, 24, 4),\n",
+ " (0, 17, 4),\n",
+ " (0, 11, 4),\n",
+ " (0, 8, 4),\n",
+ " (0, 3, 4),\n",
+ " (0, 97, 4),\n",
+ " (0, 95, 2),\n",
+ " (0, 90, 3),\n",
+ " (0, 85, 3),\n",
+ " (0, 80, 4),\n",
+ " (0, 71, 4),\n",
+ " (0, 68, 2),\n",
+ " (0, 68, 1),\n",
+ " (0, 64, 4),\n",
+ " (0, 64, 3),\n",
+ " (0, 56, 4),\n",
+ " (0, 54, 4),\n",
+ " (1, 6, 2),\n",
+ " (1, 2, 2),\n",
+ " (1, 1, 2),\n",
+ " (1, 0, 2),\n",
+ " (1, 102, 2),\n",
+ " (1, 101, 0),\n",
+ " (1, 100, 4),\n",
+ " (1, 93, 2),\n",
+ " (1, 90, 3),\n",
+ " (1, 90, 0),\n",
+ " (1, 88, 2),\n",
+ " (1, 75, 2),\n",
+ " (1, 74, 4),\n",
+ " (1, 68, 1),\n",
+ " (1, 59, 4),\n",
+ " (1, 57, 2),\n",
+ " (1, 55, 4),\n",
+ " (1, 53, 4)]"
+ ]
+ },
+ "execution_count": 16,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "%%sql\n",
+ "DROP TABLE IF EXISTS helper_output_table;\n",
+ "\n",
+ "SELECT madlib.lda_get_word_topic_mapping('outdata_predict', -- Output
table from prediction\n",
+ " 'helper_output_table');\n",
+ "\n",
+ "SELECT * FROM helper_output_table ORDER BY docid LIMIT 40;"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "metadata": {},
+ "source": [
+ "# 8. Perplexity\n",
+ "\n",
+ "Call perplexity function to see how well the model fits the data.
Perplexity computes word likelihoods averaged over the test documents."
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 17,
+ "metadata": {},
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "1 rows affected.\n"
+ ]
+ },
+ {
+ "data": {
+ "text/html": [
+ "<table>\n",
+ " <tr>\n",
+ " <th>lda_get_perplexity</th>\n",
+ " </tr>\n",
+ " <tr>\n",
+ " <td>86.6029912205</td>\n",
+ " </tr>\n",
+ "</table>"
+ ],
+ "text/plain": [
+ "[(86.6029912205131,)]"
+ ]
+ },
+ "execution_count": 17,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
+ "source": [
+ "%%sql\n",
+ "SELECT madlib.lda_get_perplexity( 'lda_model',\n",
+ " 'outdata_predict'\n",
+ " );"
+ ]
+ }
+ ],
+ "metadata": {
+ "kernelspec": {
+ "display_name": "Python 2",
+ "language": "python",
+ "name": "python2"
+ },
+ "language_info": {
+ "codemirror_mode": {
+ "name": "ipython",
+ "version": 2
+ },
+ "file_extension": ".py",
+ "mimetype": "text/x-python",
+ "name": "python",
+ "nbconvert_exporter": "python",
+ "pygments_lexer": "ipython2",
+ "version": "2.7.12"
+ }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 1
+}
